2.23

filer: leveldb and hbase may miss files when listing large directories more than 1024
fix https://github.com/chrislusf/seaweedfs/issues/1768
2026-10-01 20:26:27 +00:00 · 2021-01-31 20:18:06 -08:00 · 2021-01-31 20:11:44 -08:00 · 2021-01-31 18:26:26 -08:00 · 2021-01-30 13:38:44 -08:00 · 2021-01-30 06:18:41 -08:00
249 changed files with 8041 additions and 3245 deletions
@@ -9,6 +9,7 @@ assignees: ''

 Sponsors SeaweedFS via Patreon https://www.patreon.com/seaweedfs
 Report issues here. Ask questions here https://stackoverflow.com/questions/tagged/seaweedfs
+Please ask questions in https://github.com/chrislusf/seaweedfs/discussions

 example of a good issue report:
 https://github.com/chrislusf/seaweedfs/issues/1005
@@ -0,0 +1,22 @@
+name: Cleanup
+
+on:
+  push:
+    branches: [ master ]
+
+jobs:
+
+  build:
+    name: Build
+    runs-on: ubuntu-latest
+
+    steps:
+
+    - name: Delete old release assets
+      uses: mknejp/delete-release-assets@v1
+      with:
+        token: ${{ github.token }}
+        tag: dev
+        fail-if-no-assets: false 
+        assets: | 
+          weed-*
@@ -9,17 +9,35 @@ jobs:
  build:
    name: Build
    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        goos: [linux, windows, darwin, freebsd ]
+        goarch: [amd64, arm]
+        exclude:  
+          - goarch: arm
+            goos: darwin
+          - goarch: arm
+            goos: windows
+
    steps:

    - name: Check out code into the Go module directory
      uses: actions/checkout@v2

+    - name: Wait for the deletion
+      uses: jakejarvis/wait-action@master
+      with:
+        time: '30s'
+
+    - name: Set BUILD_TIME env
+      run: echo BUILD_TIME=$(date -u +%Y-%m-%d-%H-%M) >> ${GITHUB_ENV}
+
    - name: Go Release Binaries
-      uses: wangyoucao577/go-release-action@v1.10
+      uses: wangyoucao577/go-release-action@v1.14
      with:
        github_token: ${{ secrets.GITHUB_TOKEN }}
-        goos: linux # default is 
-        goarch: amd64 # default is 
+        goos: ${{ matrix.goos }}
+        goarch: ${{ matrix.goarch }}
        release_tag: dev
        overwrite: true 
        build_flags: -tags 5BytesOffset # optional, default is 
@@ -27,15 +45,18 @@ jobs:
        # Where to run `go build .`
        project_path: weed
        binary_name: weed-large-disk
+        asset_name: "weed-large-disk-${{ env.BUILD_TIME }}-${{ matrix.goos }}-${{ matrix.goarch }}"
+
    - name: Go Release Binaries
-      uses: wangyoucao577/go-release-action@v1.10
+      uses: wangyoucao577/go-release-action@v1.14
      with:
        github_token: ${{ secrets.GITHUB_TOKEN }}
-        goos: linux # default is 
-        goarch: amd64 # default is 
+        goos: ${{ matrix.goos }}
+        goarch: ${{ matrix.goarch }}
        release_tag: dev
        overwrite: true 
        ldflags: -extldflags -static -X github.com/chrislusf/seaweedfs/weed/util.COMMIT=${{github.sha}}
        # Where to run `go build .`
        project_path: weed
-        binary_name: weed
+        binary_name: weed-
+        asset_name: "weed-${{ env.BUILD_TIME }}-${{ matrix.goos }}-${{ matrix.goarch }}"
@@ -1,14 +1,17 @@
 # SeaweedFS


+[![Slack](https://img.shields.io/badge/slack-purple)](https://join.slack.com/t/seaweedfs/shared_invite/enQtMzI4MTMwMjU2MzA3LTEyYzZmZWYzOGQ3MDJlZWMzYmI0OTE4OTJiZjJjODBmMzUxNmYwODg0YjY3MTNlMjBmZDQ1NzQ5NDJhZWI2ZmY)
+[![Twitter](https://img.shields.io/twitter/follow/seaweedfs.svg?style=social&label=Follow)](https://twitter.com/intent/follow?screen_name=seaweedfs)
 [![Build Status](https://travis-ci.org/chrislusf/seaweedfs.svg?branch=master)](https://travis-ci.org/chrislusf/seaweedfs)
 [![GoDoc](https://godoc.org/github.com/chrislusf/seaweedfs/weed?status.svg)](https://godoc.org/github.com/chrislusf/seaweedfs/weed)
 [![Wiki](https://img.shields.io/badge/docs-wiki-blue.svg)](https://github.com/chrislusf/seaweedfs/wiki)
 [![Docker Pulls](https://img.shields.io/docker/pulls/chrislusf/seaweedfs.svg?maxAge=4800)](https://hub.docker.com/r/chrislusf/seaweedfs/)

+
 ![SeaweedFS Logo](https://raw.githubusercontent.com/chrislusf/seaweedfs/master/note/seaweedfs.png)

-<h2 align="center">Supporting SeaweedFS</h2>
+<h2 align="center"><a href="https://www.patreon.com/seaweedfs">Sponsor SeaweedFS via Patreon</a></h2>

 SeaweedFS is an independent Apache-licensed open source project with its ongoing development made
 possible entirely thanks to the support of these awesome [backers](https://github.com/chrislusf/seaweedfs/blob/master/backers.md).
@@ -17,8 +20,6 @@ If you'd like to grow SeaweedFS even stronger, please consider joining our

 Your support will be really appreciated by me and other supporters!

-<h3 align="center"><a href="https://www.patreon.com/seaweedfs">Sponsor SeaweedFS via Patreon</a></h3>
-
 <!--
 <h4 align="center">Platinum</h4>

@@ -27,24 +28,12 @@ Your support will be really appreciated by me and other supporters!
    Add your name or icon here
  </a>
 </p>
-
-<h4 align="center">Gold</h4>
-
-<table>
-  <tbody>
-    <tr>
-      <td align="center" valign="middle">
-        <a href="" target="_blank">
-          Add your name or icon here
-        </a>
-      </td>
-    </tr>
-    <tr></tr>
-  </tbody>
-</table>
-
 -->

+
+### Gold Sponsors
+![shuguang](https://raw.githubusercontent.com/chrislusf/seaweedfs/master/note/shuguang.png)
+
 ---


@@ -81,7 +70,7 @@ Table of Contents
 * Download the latest binary from https://github.com/chrislusf/seaweedfs/releases and unzip a single binary file `weed` or `weed.exe`
 * Run `weed server -dir=/some/data/dir -s3` to start one master, one volume server, one filer, and one S3 gateway.

-Also, to increase capacity, just add more volume servers by `weed volume -dir="/some/data/dir2" -mserver="<master_host>:9333" -port=8081` locally or a different machine. That is it!
+Also, to increase capacity, just add more volume servers by running `weed volume -dir="/some/data/dir2" -mserver="<master_host>:9333" -port=8081` locally, or on a different machine, or on thoudsands of machines. That is it!

 ## Introduction ##

@@ -90,15 +79,17 @@ SeaweedFS is a simple and highly scalable distributed file system. There are two
 1. to store billions of files!
 2. to serve the files fast!

-SeaweedFS started as an Object Store to handle small files efficiently. Instead of managing all file metadata in a central master, the central master only manages file volumes, and it lets these volume servers manage files and their metadata. This relieves concurrency pressure from the central master and spreads file metadata into volume servers, allowing faster file access (O(1), usually just one disk read operation).
+SeaweedFS started as an Object Store to handle small files efficiently. Instead of managing all file metadata in a central master, the central master only manages volumes on volume servers, and these volume servers manage files and their metadata. This relieves concurrency pressure from the central master and spreads file metadata into volume servers, allowing faster file access (O(1), usually just one disk read operation).

-SeaweedFS can transparently integrate with the cloud. With hot data on local cluster, and warm data on the cloud with O(1) access time, SeaweedFS can achieve both fast local access time and elastic cloud storage capacity, without any client side changes.
+SeaweedFS can transparently integrate with the cloud. With hot data on local cluster, and warm data on the cloud with O(1) access time, SeaweedFS can achieve both fast local access time and elastic cloud storage capacity. What's more, the cloud storage access API cost is minimized. Faster and Cheaper than direct cloud storage!

 There is only 40 bytes of disk storage overhead for each file's metadata. It is so simple with O(1) disk reads that you are welcome to challenge the performance with your actual use cases.

 SeaweedFS started by implementing [Facebook's Haystack design paper](http://www.usenix.org/event/osdi10/tech/full_papers/Beaver.pdf). Also, SeaweedFS implements erasure coding with ideas from [f4: Facebook’s Warm BLOB Storage System](https://www.usenix.org/system/files/conference/osdi14/osdi14-paper-muralidhar.pdf)

-On top of the object store, optional [Filer] can support directories and POSIX attributes. Filer is a separate linearly-scalable stateless server with customizable metadata stores, e.g., MySql, Postgres, Mongodb, Redis, Cassandra, Elastic Search, LevelDB, MemSql, TiDB, Etcd, CockroachDB, etc.
+On top of the object store, optional [Filer] can support directories and POSIX attributes. Filer is a separate linearly-scalable stateless server with customizable metadata stores, e.g., MySql, Postgres, Redis, Cassandra, HBase, Mongodb, Elastic Search, LevelDB, RocksDB, MemSql, TiDB, Etcd, CockroachDB, etc.
+
+For any distributed key value stores, the large values can be offloaded to SeaweedFS. With the fast access speed and linearly scalable capacity, SeaweedFS can work as a distributed [Key-Large-Value store][KeyLargeValueStore].

 [Back to TOC](#table-of-contents)

@@ -131,7 +122,10 @@ On top of the object store, optional [Filer] can support directories and POSIX a
 * [WebDAV] accesses as a mapped drive on Mac and Windows, or from mobile devices.
 * [AES256-GCM Encrypted Storage][FilerDataEncryption] safely stores the encrypted data.
 * [Super Large Files][SuperLargeFiles] stores large or super large files in tens of TB.
+
+## Kubernetes ##
 * [Kubernetes CSI Driver][SeaweedFsCsiDriver] A Container Storage Interface (CSI) Driver. [![Docker Pulls](https://img.shields.io/docker/pulls/chrislusf/seaweedfs-csi-driver.svg?maxAge=4800)](https://hub.docker.com/r/chrislusf/seaweedfs-csi-driver/)
+* [SeaweedFS Operator](https://github.com/seaweedfs/seaweedfs-operator)

 [Filer]: https://github.com/chrislusf/seaweedfs/wiki/Directories-and-Files
 [SuperLargeFiles]: https://github.com/chrislusf/seaweedfs/wiki/Data-Structure-for-Large-Files
@@ -148,6 +142,7 @@ On top of the object store, optional [Filer] can support directories and POSIX a
 [SeaweedFsCsiDriver]: https://github.com/seaweedfs/seaweedfs-csi-driver
 [ActiveActiveAsyncReplication]: https://github.com/chrislusf/seaweedfs/wiki/Filer-Active-Active-cross-cluster-continuous-synchronization
 [FilerStoreReplication]: https://github.com/chrislusf/seaweedfs/wiki/Filer-Store-Replication
+[KeyLargeValueStore]: https://github.com/chrislusf/seaweedfs/wiki/Filer-as-a-Key-Large-Value-Store

 [Back to TOC](#table-of-contents)

@@ -349,7 +344,7 @@ Usually hot data are fresh and warm data are old. SeaweedFS puts the newly creat

 With the O(1) access time, the network latency cost is kept at minimum. 

-If the hot~warm data is split as 20~80, with 20 servers, you can achieve storage capacity of 100 servers. That's a cost saving of 80%! Or you can repurpose the 80 servers to store new data also, and get 5X storage throughput.
+If the hot/warm data is split as 20/80, with 20 servers, you can achieve storage capacity of 100 servers. That's a cost saving of 80%! Or you can repurpose the 80 servers to store new data also, and get 5X storage throughput.

 [Back to TOC](#table-of-contents)

@@ -379,7 +374,7 @@ The architectures are mostly the same. SeaweedFS aims to store and read files fa

 * SeaweedFS optimizes for small files, ensuring O(1) disk seek operation, and can also handle large files.
 * SeaweedFS statically assigns a volume id for a file. Locating file content becomes just a lookup of the volume id, which can be easily cached.
-* SeaweedFS Filer metadata store can be any well-known and proven data stores, e.g., Cassandra, Mongodb, Redis, Elastic Search, MySql, Postgres, MemSql, TiDB, CockroachDB, Etcd etc, and is easy to customized.
+* SeaweedFS Filer metadata store can be any well-known and proven data stores, e.g., Redis, Cassandra, HBase, Mongodb, Elastic Search, MySql, Postgres, MemSql, TiDB, CockroachDB, Etcd etc, and is easy to customized.
 * SeaweedFS Volume server also communicates directly with clients via HTTP, supporting range queries, direct uploads, etc.

 | System         | File Metadata                   | File Content Read| POSIX  | REST API | Optimized for large number of small files |
@@ -417,11 +412,11 @@ SeaweedFS has a centralized master group to look up free volumes, while Ceph use

 Same as SeaweedFS, Ceph is also based on the object store RADOS. Ceph is rather complicated with mixed reviews.

-Ceph uses CRUSH hashing to automatically manage the data placement. SeaweedFS places data by assigned volumes.
+Ceph uses CRUSH hashing to automatically manage the data placement, which is efficient to locate the data. But the data has to be placed according to the CRUSH algorithm. Any wrong configuration would cause data loss. SeaweedFS places data by assigning them to any writable volumes. If writes to one volume failed, just pick another volume to write. Adding more volumes are also as simple as it can be.

 SeaweedFS is optimized for small files. Small files are stored as one continuous block of content, with at most 8 unused bytes between files. Small file access is O(1) disk read.

-SeaweedFS Filer uses off-the-shelf stores, such as MySql, Postgres, Mongodb, Redis, Elastic Search, Cassandra, MemSql, TiDB, CockroachCB, Etcd, to manage file directories. These stores are proven, scalable, and easier to manage.
+SeaweedFS Filer uses off-the-shelf stores, such as MySql, Postgres, Mongodb, Redis, Elastic Search, Cassandra, HBase, MemSql, TiDB, CockroachCB, Etcd, to manage file directories. These stores are proven, scalable, and easier to manage.

 | SeaweedFS         | comparable to Ceph | advantage |
 | -------------  | ------------- | ---------------- |
@@ -450,9 +445,7 @@ MinIO has specific requirements on storage layout. It is not flexible to adjust

 ## Dev Plan ##

-* More tools and documentation, on how to manage and scale the system. For example, how to move volumes, automatically balancing data, how to grow volumes, how to check system status, etc.
-* Integrate with Kubernetes. build [SeaweedFS Operator](https://github.com/seaweedfs/seaweedfs-operator).
-* Add ftp server.
+* More tools and documentation, on how to manage and scale the system.
 * Read and write stream data.
 * Support structured data.

@@ -12,6 +12,7 @@ RUN \
    SUPERCRONIC=supercronic-linux-$ARCH && \
    # Install SeaweedFS and Supercronic ( for cron job mode )
    apk add --no-cache --virtual build-dependencies --update wget curl ca-certificates && \
+    apk add fuse && \
    wget -P /tmp https://github.com/$(curl -s -L https://github.com/chrislusf/seaweedfs/releases/latest | egrep -o "chrislusf/seaweedfs/releases/download/.*/linux_$ARCH.tar.gz") && \
    tar -C /usr/bin/ -xzvf /tmp/linux_$ARCH.tar.gz && \
    curl -fsSLO "$SUPERCRONIC_URL" && \
@@ -1,7 +1,9 @@
 FROM frolvlad/alpine-glibc as builder
-RUN apk add git go g++
+RUN apk add git go g++ fuse
 RUN mkdir -p /go/src/github.com/chrislusf/
 RUN git clone https://github.com/chrislusf/seaweedfs /go/src/github.com/chrislusf/seaweedfs
+ARG BRANCH=${BRANCH:-master}
+RUN cd /go/src/github.com/chrislusf/seaweedfs && git checkout $BRANCH
 RUN cd /go/src/github.com/chrislusf/seaweedfs/weed \
  && export LDFLAGS="-X github.com/chrislusf/seaweedfs/weed/util.COMMIT=$(git rev-parse --short HEAD)" \
  && go install -ldflags "${LDFLAGS}"
@@ -12,6 +14,7 @@ COPY --from=builder /root/go/bin/weed /usr/bin/
 RUN mkdir -p /etc/seaweedfs
 COPY --from=builder /go/src/github.com/chrislusf/seaweedfs/docker/filer.toml /etc/seaweedfs/filer.toml
 COPY --from=builder /go/src/github.com/chrislusf/seaweedfs/docker/entrypoint.sh /entrypoint.sh
+RUN apk add fuse # for weed mount

 # volume server gprc port
 EXPOSE 18080
@@ -1,7 +1,9 @@
 FROM frolvlad/alpine-glibc as builder
-RUN apk add git go g++
+RUN apk add git go g++ fuse
 RUN mkdir -p /go/src/github.com/chrislusf/
 RUN git clone https://github.com/chrislusf/seaweedfs /go/src/github.com/chrislusf/seaweedfs
+ARG BRANCH=${BRANCH:-master}
+RUN cd /go/src/github.com/chrislusf/seaweedfs && git checkout $BRANCH
 RUN cd /go/src/github.com/chrislusf/seaweedfs/weed \
  && export LDFLAGS="-X github.com/chrislusf/seaweedfs/weed/util.COMMIT=$(git rev-parse --short HEAD)" \
  && go install -tags 5BytesOffset -ldflags "${LDFLAGS}"
@@ -12,6 +14,7 @@ COPY --from=builder /root/go/bin/weed /usr/bin/
 RUN mkdir -p /etc/seaweedfs
 COPY --from=builder /go/src/github.com/chrislusf/seaweedfs/docker/filer.toml /etc/seaweedfs/filer.toml
 COPY --from=builder /go/src/github.com/chrislusf/seaweedfs/docker/entrypoint.sh /entrypoint.sh
+RUN apk add fuse # for weed mount

 # volume server gprc port
 EXPOSE 18080
@@ -4,6 +4,7 @@ COPY  ./weed /usr/bin/
 RUN mkdir -p /etc/seaweedfs
 COPY ./filer.toml /etc/seaweedfs/filer.toml
 COPY ./entrypoint.sh /entrypoint.sh
+RUN apk add fuse # for weed mount

 # volume server grpc port
 EXPOSE 18080
@@ -12,11 +12,26 @@ build:
 dev: build
 	docker-compose -f local-dev-compose.yml -p seaweedfs up

+dev_mount: build
+	docker-compose -f local-mount-compose.yml -p seaweedfs up
+
+k8s: build
+	docker-compose -f local-k8s-compose.yml -p seaweedfs up
+
 dev_registry: build
 	docker-compose -f local-registry-compose.yml -p seaweedfs up

+dev_replicate: build
+	docker-compose -f local-replicate-compose.yml -p seaweedfs up
+
 cluster: build
 	docker-compose -f local-cluster-compose.yml -p seaweedfs up

+2clusters: build
+	docker-compose -f local-clusters-compose.yml -p seaweedfs up
+
+filer_etcd: build
+	docker stack deploy -c swarm-etcd.yml fs
+
 clean:
 	rm ./weed
@@ -0,0 +1,23 @@
+version: '2'
+
+services:
+  server1:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9333:9333
+      - 19333:19333
+      - 8084:8080
+      - 18084:18080
+      - 8888:8888
+      - 18888:18888
+    command: "server -ip=server1 -filer -volume.max=0 -master.volumeSizeLimitMB=1024 -volume.preStopSeconds=1"
+  server2:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9334:9333
+      - 19334:19333
+      - 8085:8080
+      - 18085:18080
+      - 8889:8888
+      - 18889:18888
+    command: "server -ip=server2 -filer -volume.max=0 -master.volumeSizeLimitMB=1024 -volume.preStopSeconds=1"
@@ -12,7 +12,7 @@ services:
    ports:
      - 8080:8080
      - 18080:18080
-    command: "volume -mserver=master:9333 -port=8080 -ip=volume"
+    command: "volume -mserver=master:9333 -port=8080 -ip=volume -preStopSeconds=1"
    depends_on:
      - master
  filer:
@@ -33,3 +33,14 @@ services:
      - master
      - volume
      - filer
+  mount:
+    image: chrislusf/seaweedfs:local
+    privileged: true
+    cap_add:
+      - SYS_ADMIN
+    mem_limit: 4096m
+    command: '-v=4 mount -filer="filer:8888" -dirAutoCreate -dir=/mnt/seaweedfs -cacheCapacityMB=100 -concurrentWriters=128'
+    depends_on:
+      - master
+      - volume
+      - filer
@@ -0,0 +1,94 @@
+version: '2'
+
+services:
+  master:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9333:9333
+      - 19333:19333
+    command: "master -ip=master"
+  volume:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8080:8080
+      - 18080:18080
+    command: "volume -mserver=master:9333 -port=8080 -ip=volume"
+    depends_on:
+      - master
+  mysql:
+    image: percona/percona-server:5.7
+    ports:
+      - 3306:3306
+    volumes:
+      - ./seaweedfs.sql:/docker-entrypoint-initdb.d/seaweedfs.sql
+    environment:
+      - MYSQL_ROOT_PASSWORD=secret
+      - MYSQL_DATABASE=seaweedfs
+      - MYSQL_PASSWORD=secret
+      - MYSQL_USER=seaweedfs
+  filer:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8888:8888
+      - 18888:18888
+    environment:
+      - WEED_MYSQL_HOSTNAME=mysql
+      - WEED_MYSQL_PORT=3306
+      - WEED_MYSQL_DATABASE=seaweedfs
+      - WEED_MYSQL_USERNAME=seaweedfs
+      - WEED_MYSQL_PASSWORD=secret
+      - WEED_MYSQL_ENABLED=true
+      - WEED_MYSQL_CONNECTION_MAX_IDLE=5
+      - WEED_MYSQL_CONNECTION_MAX_OPEN=75
+      # "refresh" connection every 10 minutes, eliminating mysql closing "old" connections
+      - WEED_MYSQL_CONNECTION_MAX_LIFETIME_SECONDS=600
+      # enable usage of memsql as filer backend
+      - WEED_MYSQL_INTERPOLATEPARAMS=true
+      - WEED_LEVELDB2_ENABLED=false
+    command: '-v 9 filer -master="master:9333"'
+    depends_on:
+      - master
+      - volume
+      - mysql
+  ingress:
+    image: jwilder/nginx-proxy:alpine
+    ports:
+      - "80:80"
+    volumes:
+      - /var/run/docker.sock:/tmp/docker.sock:ro
+      - ./nginx/proxy.conf:/etc/nginx/proxy.conf
+  s3:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8333:8333
+    command: '-v 9 s3 -filer="filer:8888"'
+    depends_on:
+      - master
+      - volume
+      - filer
+    environment:
+      - VIRTUAL_HOST=ingress
+      - VIRTUAL_PORT=8333
+  registry:
+    image: registry:2
+    environment:
+      REGISTRY_HTTP_ADDR: "0.0.0.0:5001" # seaweedfs s3
+      REGISTRY_LOG_LEVEL: "debug"
+      REGISTRY_STORAGE: "s3"
+      REGISTRY_STORAGE_S3_REGION: "us-east-1"
+      REGISTRY_STORAGE_S3_REGIONENDPOINT: "http://ingress"
+      REGISTRY_STORAGE_S3_BUCKET: "registry"
+      REGISTRY_STORAGE_S3_ACCESSKEY: "some_access_key1"
+      REGISTRY_STORAGE_S3_SECRETKEY: "some_secret_key1"
+      REGISTRY_STORAGE_S3_V4AUTH: "true"
+      REGISTRY_STORAGE_S3_SECURE: "false"
+      REGISTRY_STORAGE_S3_SKIPVERIFY: "true"
+      REGISTRY_STORAGE_S3_ROOTDIRECTORY: "/"
+      REGISTRY_STORAGE_DELETE_ENABLED: "true"
+      REGISTRY_STORAGE_REDIRECT_DISABLE: "true"
+      REGISTRY_VALIDATION_DISABLED: "true"
+    ports:
+      - 5001:5001
+    depends_on:
+      - s3
+      - ingress
@@ -0,0 +1,46 @@
+version: '2'
+
+services:
+  master:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9333:9333
+      - 19333:19333
+    command: "master -ip=master"
+  volume:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 7455:8080
+      - 9325:9325
+    command: 'volume -mserver="master:9333" -port=8080 -metricsPort=9325 -preStopSeconds=1 -publicUrl=localhost:7455'
+    depends_on:
+      - master
+  filer:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8888:8888
+      - 18888:18888
+      - 9326:9326
+    command: 'filer -master="master:9333"  -metricsPort=9326'
+    tty: true
+    stdin_open: true
+    depends_on:
+      - master
+      - volume
+  mount_1:
+    image: chrislusf/seaweedfs:local
+    privileged: true
+    entrypoint: '/bin/sh -c "mkdir -p t1 && mkdir -p cache/t1 && weed -v=4 mount -filer=filer:8888 -cacheDir=./cache/t1 -dir=./t1 -filer.path=/c1" -volumeServerAccess=filerProxy'
+    depends_on:
+      - master
+      - volume
+      - filer
+  mount_2:
+    image: chrislusf/seaweedfs:local
+    privileged: true
+    entrypoint: '/bin/sh -c "mkdir -p t2 && mkdir -p cache/t2 && weed -v=4  mount -filer=filer:8888 -cacheDir=./cache/t2 -dir=./t2 -filer.path=/c1" -volumeServerAcess=publicUrl'
+    depends_on:
+      - master
+      - volume
+      - filer
+      - mount_1
@@ -13,10 +13,6 @@ services:
      - 8080:8080
      - 18080:18080
    command: "volume -mserver=master:9333 -port=8080 -ip=volume -max=0 -preStopSeconds=1"
-    volumes:
-      - type: bind
-        source: /Volumes/mobile_disk/data
-        target: /data
    depends_on:
      - master
  filer:
@@ -0,0 +1,59 @@
+version: '2'
+
+services:
+  master:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9333:9333
+      - 19333:19333
+    command: "master -ip=master"
+  volume:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8080:8080
+      - 18080:18080
+    command: "volume -mserver=master:9333 -port=8080 -ip=volume -preStopSeconds=1"
+    depends_on:
+      - master
+  filer:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8888:8888
+      - 18888:18888
+    command: '-v=9 filer -master="master:9333"'
+    restart: on-failure
+    volumes:
+      - ./notification.toml:/etc/seaweedfs/notification.toml
+    depends_on:
+      - master
+      - volume
+      - rabbitmq
+      - replicate
+    environment:
+      RABBIT_SERVER_URL: "amqp://guest:guest@rabbitmq:5672/"
+  replicate:
+    image: chrislusf/seaweedfs:local
+    command: '-v=9 filer.replicate'
+    restart: on-failure
+    volumes:
+      - ./notification.toml:/etc/seaweedfs/notification.toml
+      - ./replication.toml:/etc/seaweedfs/replication.toml
+    depends_on:
+      - rabbitmq
+    environment:
+      RABBIT_SERVER_URL: "amqp://guest:guest@rabbitmq:5672/"
+  s3:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8333:8333
+    command: 's3 -filer="filer:8888"'
+    depends_on:
+      - master
+      - volume
+      - filer
+  rabbitmq:
+    image: rabbitmq:3.8.10-management-alpine
+    ports:
+      - 5672:5672
+      - 15671:15671
+      - 15672:15672
@@ -0,0 +1,30 @@
+# HTTP 1.1 support
+proxy_http_version 1.1;
+#proxy_buffering off;
+proxy_set_header Host $http_host;
+proxy_set_header Upgrade $http_upgrade;
+proxy_set_header Connection $proxy_connection;
+proxy_set_header X-Real-IP $remote_addr;
+proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
+proxy_set_header X-Forwarded-Proto $proxy_x_forwarded_proto;
+proxy_set_header X-Forwarded-Ssl $proxy_x_forwarded_ssl;
+proxy_set_header X-Forwarded-Port $proxy_x_forwarded_port;
+
+# Mitigate httpoxy attack (see README for details)
+proxy_set_header Proxy "";
+
+# aws default max_concurrent_requests 10
+# aws default multipart_threshold     8MB
+proxy_buffering on;     # GET buffering or “X-Accel-Buffering” enables or disables buffering of a response;
+proxy_buffers 64 1m;  # buffers used for reading a response from the proxied server, for a single connection
+proxy_buffer_size 8k; # maximum size of the data that nginx can receive from the server at a time is set
+proxy_busy_buffers_size 2m;
+
+proxy_request_buffering on;  # PUT buffering
+client_body_buffer_size 64m; # buffer size for reading client request body
+client_max_body_size    64m;
+
+proxy_next_upstream error timeout non_idempotent http_500; # PUT request should be passed to the next server:
+proxy_connect_timeout 200ms;
+proxy_read_timeout    3s; #timeout is set only between two successive read operations
+proxy_send_timeout    3s; #timeout is set only between two successive write operations
@@ -0,0 +1,17 @@
+[notification.log]
+# this is only for debugging perpose and does not work with "weed filer.replicate"
+enabled = false
+
+
+[notification.gocdk_pub_sub]
+# The Go Cloud Development Kit (https://gocloud.dev).
+# PubSub API (https://godoc.org/gocloud.dev/pubsub).
+# Supports AWS SNS/SQS, Azure Service Bus, Google PubSub, NATS and RabbitMQ.
+enabled = true
+# This URL will Dial the RabbitMQ server at the URL in the environment
+# variable RABBIT_SERVER_URL and open the exchange "myexchange".
+# The exchange must have already been created by some other means, like
+# the RabbitMQ management plugin. Сreate myexchange of type fanout and myqueue then
+# create binding myexchange => myqueue
+topic_url = "rabbit://swexchange"
+sub_url = "rabbit://swqueue"
@@ -0,0 +1,13 @@
+global:
+  scrape_interval: 30s
+  scrape_timeout: 10s
+
+scrape_configs:
+  - job_name: services
+    metrics_path: /metrics
+    static_configs:
+      - targets:
+          - 'prometheus:9090'
+          - 'volume:9325'
+          - 'filer:9326'
+          - 's3:9327'
@@ -0,0 +1,11 @@
+[source.filer]
+enabled = true
+grpcAddress = "filer:18888"
+# all files under this directory tree are replicated.
+# this is not a directory on your hard drive, but on your filer.
+# i.e., all files with this "prefix" are sent to notification message queue.
+directory = "/buckets"
+
+[sink.local_incremental]
+enabled = true
+directory = "/data"
@@ -12,7 +12,8 @@ services:
    ports:
      - 8080:8080
      - 18080:18080
-    command: 'volume -mserver="master:9333" -port=8080'
+      - 9325:9325
+    command: 'volume -mserver="master:9333" -port=8080  -metricsPort=9325'
    depends_on:
      - master
  filer:
@@ -20,7 +21,8 @@ services:
    ports:
      - 8888:8888
      - 18888:18888
-    command: 'filer -master="master:9333"'
+      - 9326:9326
+    command: 'filer -master="master:9333"  -metricsPort=9326'
    tty: true
    stdin_open: true
    depends_on:
@@ -40,8 +42,18 @@ services:
    image: chrislusf/seaweedfs # use a remote image
    ports:
      - 8333:8333
-    command: 's3 -filer="filer:8888"'
+      - 9327:9327
+    command: 's3 -filer="filer:8888" -metricsPort=9327'
    depends_on:
      - master
      - volume
      - filer
+  prometheus:
+    image: prom/prometheus:v2.21.0
+    ports:
+      - 9000:9090
+    volumes:
+      - ./prometheus:/etc/prometheus
+    command: --web.enable-lifecycle  --config.file=/etc/prometheus/prometheus.yml
+    depends_on:
+      - s3
@@ -0,0 +1,12 @@
+CREATE DATABASE IF NOT EXISTS seaweedfs;
+CREATE USER IF NOT EXISTS 'seaweedfs'@'%' IDENTIFIED BY 'secret';
+GRANT ALL PRIVILEGES ON seaweedfs_fast.* TO 'seaweedfs'@'%';
+FLUSH PRIVILEGES;
+USE seaweedfs;
+CREATE TABLE IF NOT EXISTS filemeta (
+    dirhash     BIGINT         COMMENT 'first 64 bits of MD5 hash value of directory field',
+    name        VARCHAR(1000)  COMMENT 'directory or file name',
+    directory   TEXT           COMMENT 'full path to parent directory',
+    meta        LONGBLOB,
+    PRIMARY KEY (dirhash, name)
+) DEFAULT CHARSET=utf8;
@@ -0,0 +1,84 @@
+# 2021-01-30 16:25:30
+version: '3.8'
+
+services:
+
+  etcd:
+    image: gasparekatapy/etcd
+    networks:
+      - net
+    deploy:
+      mode: replicated
+      replicas: 3
+
+  master:
+    image: chrislusf/seaweedfs:local
+    environment:
+      WEED_MASTER_FILER_DEFAULT: "filer:8888"
+      WEED_MASTER_SEQUENCER_TYPE: "raft"
+    ports:
+      - "9333:9333"
+      - "19333:19333"
+    networks:
+      - net
+    command:
+      - 'master'
+      - '-resumeState=true'
+      - '-ip=master'
+      - '-port=9333'
+    deploy:
+      mode: replicated
+      replicas: 1
+
+  filer:
+    image: chrislusf/seaweedfs:local
+    environment:
+      WEED_LEVELDB2_ENABLED: "false"
+      WEED_ETCD_ENABLED: "true"
+      WEED_ETCD_SERVERS: "etcd:2379"
+    ports:
+      - target: 8888
+        published: 8888
+        protocol: tcp
+        mode: host
+      - target: 18888
+        published: 18888
+        protocol: tcp
+        mode: host
+    networks:
+      - net
+    command:
+      - 'filer'
+      - '-ip=filer'
+      - '-port=8888'
+      - '-port.readonly=28888'
+      - '-master=master:9333'
+      - '-disableDirListing=true'
+    deploy:
+      mode: replicated
+      replicas: 1
+
+  volume:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - target: 8080
+        published: 8080
+        protocol: tcp
+        mode: host
+      - target: 18080
+        published: 18080
+        protocol: tcp
+        mode: host
+    networks:
+      - net
+    command:
+      - 'volume'
+      - '-mserver=master:9333'
+      - '-port=8080'
+    deploy:
+      mode: global
+
+  ###########################################################################
+
+networks:
+  net:
@@ -8,10 +8,10 @@ require (
 	github.com/Azure/azure-storage-blob-go v0.8.0
 	github.com/OneOfOne/xxhash v1.2.2
 	github.com/Shopify/sarama v1.23.1
-	github.com/aws/aws-sdk-go v1.33.5
+	github.com/aws/aws-sdk-go v1.34.30
 	github.com/buraksezer/consistent v0.0.0-20191006190839-693edf70fd72
 	github.com/cespare/xxhash v1.1.0
-	github.com/chrislusf/raft v1.0.3
+	github.com/chrislusf/raft v1.0.4
 	github.com/coreos/go-semver v0.3.0 // indirect
 	github.com/dgrijalva/jwt-go v3.2.0+incompatible
 	github.com/disintegration/imaging v1.6.2
@@ -25,7 +25,7 @@ require (
 	github.com/fclairamb/ftpserverlib v0.8.0
 	github.com/frankban/quicktest v1.7.2 // indirect
 	github.com/go-errors/errors v1.1.1 // indirect
-	github.com/go-redis/redis v6.15.7+incompatible
+	github.com/go-redis/redis/v8 v8.4.4
 	github.com/go-sql-driver/mysql v1.5.0
 	github.com/gocql/gocql v0.0.0-20190829130954-e163eff7a8c6
 	github.com/gogo/protobuf v1.2.2-0.20190730201129-28a6bbf47e48 // indirect
@@ -39,7 +39,7 @@ require (
 	github.com/jcmturner/gofork v1.0.0 // indirect
 	github.com/json-iterator/go v1.1.10
 	github.com/karlseguin/ccache v2.0.3+incompatible
-	github.com/karlseguin/expect v1.0.1 // indirect
+	github.com/karlseguin/ccache/v2 v2.0.7
 	github.com/klauspost/compress v1.10.9 // indirect
 	github.com/klauspost/cpuid v1.2.1 // indirect
 	github.com/klauspost/crc32 v1.2.0
@@ -51,14 +51,12 @@ require (
 	github.com/mattn/go-ieproxy v0.0.0-20190805055040-f9202b1cfdeb // indirect
 	github.com/mattn/go-runewidth v0.0.4 // indirect
 	github.com/olivere/elastic/v7 v7.0.19
-	github.com/onsi/ginkgo v1.10.1 // indirect
-	github.com/onsi/gomega v1.7.0 // indirect
 	github.com/peterh/liner v1.1.0
 	github.com/pierrec/lz4 v2.2.7+incompatible // indirect
 	github.com/prometheus/client_golang v1.3.0
 	github.com/rakyll/statik v0.1.7
 	github.com/rcrowley/go-metrics v0.0.0-20190826022208-cac0b30c2563 // indirect
-	github.com/seaweedfs/fuse v1.0.7
+	github.com/seaweedfs/fuse v1.1.1
 	github.com/seaweedfs/goexif v1.0.2
 	github.com/skip2/go-qrcode v0.0.0-20200617195104-da1b6568686e
 	github.com/spaolacci/murmur3 v1.1.0 // indirect
@@ -66,8 +64,10 @@ require (
 	github.com/spf13/viper v1.4.0
 	github.com/stretchr/testify v1.6.1
 	github.com/syndtr/goleveldb v1.0.0
+	github.com/tecbot/gorocksdb v0.0.0-20191217155057-f0fad39f321c
 	github.com/tidwall/gjson v1.3.2
 	github.com/tidwall/match v1.0.1
+	github.com/tsuna/gohbase v0.0.0-20201125011725-348991136365
 	github.com/valyala/bytebufferpool v1.0.0
 	github.com/viant/assertly v0.5.4 // indirect
 	github.com/viant/ptrie v0.3.0
@@ -81,7 +81,7 @@ require (
 	gocloud.dev/pubsub/natspubsub v0.16.0
 	gocloud.dev/pubsub/rabbitpubsub v0.16.0
 	golang.org/x/image v0.0.0-20200119044424-58c23975cae1 // indirect
-	golang.org/x/net v0.0.0-20200202094626-16171245cfb2
+	golang.org/x/net v0.0.0-20201202161906-c7110b5ffcbb
 	golang.org/x/sync v0.0.0-20200930132711-30421366ff76 // indirect
 	golang.org/x/sys v0.0.0-20201022201747-fb209a7c41cd
 	golang.org/x/tools v0.0.0-20200103221440-774c71fcf114
@@ -61,6 +61,8 @@ github.com/aws/aws-sdk-go v1.19.45/go.mod h1:KmX6BPdI08NWTb3/sm4ZGu5ShLoqVDhKgpi
 github.com/aws/aws-sdk-go v1.27.0/go.mod h1:KmX6BPdI08NWTb3/sm4ZGu5ShLoqVDhKgpiN924inxo=
 github.com/aws/aws-sdk-go v1.33.5 h1:p2fr1ryvNTU6avUWLI+/H7FGv0TBIjzVM5WDgXBBv4U=
 github.com/aws/aws-sdk-go v1.33.5/go.mod h1:5zCpMtNQVjRREroY7sYe8lOMRSxkhG6MZveU8YkpAk0=
+github.com/aws/aws-sdk-go v1.34.30 h1:izATc/E0+HcT5YHmaQVjn7GHCoqaBxn0PGo6Zq5UNFA=
+github.com/aws/aws-sdk-go v1.34.30/go.mod h1:H7NKnBqNVzoTJpGfLrQkkD+ytBA93eiDYi/+8rV9s48=
 github.com/aws/aws-sdk-go-v2 v0.18.0/go.mod h1:JWVYvqSMppoMJC0x5wdwiImzgXTI9FuZwxzkQq9wy+g=
 github.com/beorn7/perks v0.0.0-20180321164747-3a771d992973/go.mod h1:Dwedo/Wpr24TaqPxmxbtue+5NUziq4I4S80YR8gNf3Q=
 github.com/beorn7/perks v1.0.0 h1:HWo1m869IqiPhD389kmkxeTalrjNbbJTC8LXupb+sl0=
@@ -84,8 +86,8 @@ github.com/cespare/xxhash v1.1.0 h1:a6HrQnmkObjyL+Gs60czilIUGqrzKutQD6XZog3p+ko=
 github.com/cespare/xxhash v1.1.0/go.mod h1:XrSqR1VqqWfGrhpAt58auRo0WTKS1nRRg3ghfAqPWnc=
 github.com/cespare/xxhash/v2 v2.1.1 h1:6MnRN8NT7+YBpUIWxHtefFZOKTAPgGjpQSxqLNn0+qY=
 github.com/cespare/xxhash/v2 v2.1.1/go.mod h1:VGX0DQ3Q6kWi7AoAeZDth3/j3BFtOZR5XLFGgcrjCOs=
-github.com/chrislusf/raft v1.0.3 h1:11YrnzJtVa5z7m9lhY2p8VcPHoUlC1UswyoAo+U1m1k=
-github.com/chrislusf/raft v1.0.3/go.mod h1:Ep5DP+mJSosjfKiix1uU7Lc2Df/SX4oGJEpZlXH5l68=
+github.com/chrislusf/raft v1.0.4 h1:THhbsVik2hxdE0/VXX834f64Wn9RzgVPp+E+XCWZdKM=
+github.com/chrislusf/raft v1.0.4/go.mod h1:Ep5DP+mJSosjfKiix1uU7Lc2Df/SX4oGJEpZlXH5l68=
 github.com/clbanning/x2j v0.0.0-20191024224557-825249438eec/go.mod h1:jMjuTZXRI4dUb/I5gc9Hdhagfvm9+RyrPryS/auMzxE=
 github.com/client9/misspell v0.3.4/go.mod h1:qj6jICC3Q7zFZvVWo7KLAzC3yx5G7kyvSDkc90ppPyw=
 github.com/cncf/udpa/go v0.0.0-20191209042840-269d4d468f6f/go.mod h1:M8M6+tZqaGXZJjfX53e64911xZQV5JYwmTeXPW+k8Sc=
@@ -114,6 +116,8 @@ github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSs
 github.com/devigned/tab v0.1.1/go.mod h1:XG9mPq0dFghrYvoBF3xdRrJzSTX1b7IQrvaL9mzjeJY=
 github.com/dgrijalva/jwt-go v3.2.0+incompatible h1:7qlOGliEKZXTDg6OTjfoBKDXWrumCAMpl/TFQ4/5kLM=
 github.com/dgrijalva/jwt-go v3.2.0+incompatible/go.mod h1:E3ru+11k8xSBh+hMPgOLZmtrrCbhqsmaPHjLKYnJCaQ=
+github.com/dgryski/go-rendezvous v0.0.0-20200823014737-9f7001d12a5f h1:lO4WD4F/rVNCu3HqELle0jiPLLBs70cWOduZpkS1E78=
+github.com/dgryski/go-rendezvous v0.0.0-20200823014737-9f7001d12a5f/go.mod h1:cuUVRXasLTGF7a8hSLbxyZXjz+1KgoB3wDUb6vlszIc=
 github.com/dgryski/go-sip13 v0.0.0-20181026042036-e10d5fee7954/go.mod h1:vAd38F8PWV+bWy6jNmig1y/TA+kYO4g3RSRF0IAv0no=
 github.com/dimchansky/utfbom v1.1.0/go.mod h1:rO41eb7gLfo8SF1jd9F8HplJm1Fewwi4mQvIirEdv+8=
 github.com/disintegration/imaging v1.6.2 h1:w1LecBlG2Lnp8B3jk5zSuNqd7b4DXhcjwek1ei82L+c=
@@ -157,6 +161,8 @@ github.com/frankban/quicktest v1.7.2 h1:2QxQoC1TS09S7fhCPsrvqYdvP1H5M1P1ih5ABm3B
 github.com/frankban/quicktest v1.7.2/go.mod h1:jaStnuzAqU1AJdCO0l53JDCJrVDKcS03DbaAcR7Ks/o=
 github.com/fsnotify/fsnotify v1.4.7 h1:IXs+QLmnXW2CcXuY+8Mzv/fWEsPGWxqefPtCP5CnV9I=
 github.com/fsnotify/fsnotify v1.4.7/go.mod h1:jwhsz4b93w/PPRr/qN1Yymfu8t87LnFCMoQvtojpjFo=
+github.com/fsnotify/fsnotify v1.4.9 h1:hsms1Qyu0jgnwNXIxa+/V/PDsU6CfLf6CNO8H7IWoS4=
+github.com/fsnotify/fsnotify v1.4.9/go.mod h1:znqG4EE+3YCdAaPaxE2ZRY/06pZUdp0tY4IgpuI1SZQ=
 github.com/ghodss/yaml v1.0.0 h1:wQHKEahhL6wmXdzwWG11gIVCkOv05bNOh+Rxn0yngAk=
 github.com/ghodss/yaml v1.0.0/go.mod h1:4dBDuWmgqj2HViK6kFavaiC9ZROes6MMH2rRYeMEF04=
 github.com/go-errors/errors v1.1.1 h1:ljK/pL5ltg3qoN+OtN6yCv9HWSfMwxSx90GJCZQxYNg=
@@ -170,8 +176,8 @@ github.com/go-logfmt/logfmt v0.3.0/go.mod h1:Qt1PoO58o5twSAckw1HlFXLmHsOX5/0LbT9
 github.com/go-logfmt/logfmt v0.4.0/go.mod h1:3RMwSq7FuexP4Kalkev3ejPJsZTpXXBr9+V4qmtdjCk=
 github.com/go-logfmt/logfmt v0.5.0 h1:TrB8swr/68K7m9CcGut2g3UOihhbcbiMAYiuTXdEih4=
 github.com/go-logfmt/logfmt v0.5.0/go.mod h1:wCYkCAKZfumFQihp8CzCvQ3paCTfi41vtzG1KdI/P7A=
-github.com/go-redis/redis v6.15.7+incompatible h1:3skhDh95XQMpnqeqNftPkQD9jL9e5e36z/1SUm6dy1U=
-github.com/go-redis/redis v6.15.7+incompatible/go.mod h1:NAIEuMOZ/fxfXJIrKDQDz8wamY7mA7PouImQ2Jvg6kA=
+github.com/go-redis/redis/v8 v8.4.4 h1:fGqgxCTR1sydaKI00oQf3OmkU/DIe/I/fYXvGklCIuc=
+github.com/go-redis/redis/v8 v8.4.4/go.mod h1:nA0bQuF0i5JFx4Ta9RZxGKXFrQ8cRWntra97f0196iY=
 github.com/go-sql-driver/mysql v1.4.0/go.mod h1:zAC/RDZ24gD3HViQzih4MyKcchzm+sOG5ZlKdlhCg5w=
 github.com/go-sql-driver/mysql v1.4.1 h1:g24URVg0OFbNUTx9qqY1IRZ9D9z3iPyi5zKhQZpNwpA=
 github.com/go-sql-driver/mysql v1.4.1/go.mod h1:zAC/RDZ24gD3HViQzih4MyKcchzm+sOG5ZlKdlhCg5w=
@@ -179,6 +185,8 @@ github.com/go-sql-driver/mysql v1.5.0 h1:ozyZYNQW3x3HtqT1jira07DN2PArx2v7/mN66gG
 github.com/go-sql-driver/mysql v1.5.0/go.mod h1:DCzpHaOWr8IXmIStZouvnhqoel9Qv2LBy8hT2VhHyBg=
 github.com/go-stack/stack v1.8.0 h1:5SgMzNM5HxrEjV0ww2lTmX6E2Izsfxas4+YHWRs3Lsk=
 github.com/go-stack/stack v1.8.0/go.mod h1:v0f6uXyyMGvRgIKkXu+yp6POWl0qKG85gN/melR3HDY=
+github.com/go-zookeeper/zk v1.0.2 h1:4mx0EYENAdX/B/rbunjlt5+4RTA/a9SMHBRuSKdGxPM=
+github.com/go-zookeeper/zk v1.0.2/go.mod h1:nOB03cncLtlp4t+UAkGSV+9beXP/akpekBwL+UX1Qcw=
 github.com/gobuffalo/attrs v0.0.0-20190224210810-a9411de4debd/go.mod h1:4duuawTqi2wkkpB4ePgWMaai6/Kc6WEz83bhFwpHzj0=
 github.com/gobuffalo/depgen v0.0.0-20190329151759-d478694a28d3/go.mod h1:3STtPUQYuzV0gBVOY3vy6CfMm/ljR4pABfrTeHNLHUY=
 github.com/gobuffalo/depgen v0.1.0/go.mod h1:+ifsuy7fhi15RWncXQQKjWS9JPkdah5sZvtHc2RXGlg=
@@ -223,6 +231,8 @@ github.com/golang/groupcache v0.0.0-20190702054246-869f871628b6/go.mod h1:cIg4er
 github.com/golang/mock v1.1.1/go.mod h1:oTYuIxOrZwtPieC+H1uAHpcLFnEyAGVDL/k47Jfbm0A=
 github.com/golang/mock v1.2.0/go.mod h1:oTYuIxOrZwtPieC+H1uAHpcLFnEyAGVDL/k47Jfbm0A=
 github.com/golang/mock v1.3.1/go.mod h1:sBzyDLLjw3U8JLTeZvSv8jJB+tU5PVekmnlKIyFUx0Y=
+github.com/golang/mock v1.4.3 h1:GV+pQPG/EUUbkh47niozDcADz6go/dUwhVzdUQHIVRw=
+github.com/golang/mock v1.4.3/go.mod h1:UOMv5ysSaYNkG+OFQykRIcU/QvvxJf3p21QfJ2Bt3cw=
 github.com/golang/protobuf v1.2.0/go.mod h1:6lQm79b+lXiMfvg/cZm0SGofjICqVBUtrP5yJMmIC1U=
 github.com/golang/protobuf v1.3.1/go.mod h1:6lQm79b+lXiMfvg/cZm0SGofjICqVBUtrP5yJMmIC1U=
 github.com/golang/protobuf v1.3.2 h1:6nsPYzhq5kReh6QImI3k5qWzO4PEbvbIW2cwSfR/6xs=
@@ -253,6 +263,8 @@ github.com/google/go-cmp v0.4.0 h1:xsAVV57WRhGj6kEIi8ReJzQlHHqcBYCElAvkovg3B/4=
 github.com/google/go-cmp v0.4.0/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
 github.com/google/go-cmp v0.5.0 h1:/QaMHBdZ26BB3SSst0Iwl10Epc+xhTquomWX0oZEB6w=
 github.com/google/go-cmp v0.5.0/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
+github.com/google/go-cmp v0.5.4 h1:L8R9j+yAqZuZjsqh/z+F1NCffTKKLShY6zXTItVIZ8M=
+github.com/google/go-cmp v0.5.4/go.mod h1:v8dTdLbMG2kIc/vJvl+f65V22dbkXbowE6jgT/gNBxE=
 github.com/google/go-replayers/grpcreplay v0.1.0 h1:eNb1y9rZFmY4ax45uEEECSa8fsxGRU+8Bil52ASAwic=
 github.com/google/go-replayers/grpcreplay v0.1.0/go.mod h1:8Ig2Idjpr6gifRd6pNVggX6TC1Zw6Jx74AKp7QNH2QE=
 github.com/google/go-replayers/httpreplay v0.1.0 h1:AX7FUb4BjrrzNvblr/OlgwrmFiep6soj5K2QSDW7BGk=
@@ -339,6 +351,10 @@ github.com/jmespath/go-jmespath v0.0.0-20180206201540-c2b33e8439af h1:pmfjZENx5i
 github.com/jmespath/go-jmespath v0.0.0-20180206201540-c2b33e8439af/go.mod h1:Nht3zPeWKUH0NzdCt2Blrr5ys8VGpn0CEB0cQHVjt7k=
 github.com/jmespath/go-jmespath v0.3.0 h1:OS12ieG61fsCg5+qLJ+SsW9NicxNkg3b25OyT2yCeUc=
 github.com/jmespath/go-jmespath v0.3.0/go.mod h1:9QtRXoHjLGCJ5IBSaohpXITPlowMeeYCZ7fLUTSywik=
+github.com/jmespath/go-jmespath v0.4.0 h1:BEgLn5cpjn8UN1mAw4NjwDrS35OdebyEtFe+9YPoQUg=
+github.com/jmespath/go-jmespath v0.4.0/go.mod h1:T8mJZnbsbmF+m6zOOFylbeCJqk5+pHWvzYPziyZiYoo=
+github.com/jmespath/go-jmespath/internal/testify v1.5.1 h1:shLQSRRSCCPj3f2gpwzGwWFoC7ycTf1rcQZHOlsJ6N8=
+github.com/jmespath/go-jmespath/internal/testify v1.5.1/go.mod h1:L3OGu8Wl2/fWfCI6z80xFu9LTZmf1ZRjMHUOPmWr69U=
 github.com/joho/godotenv v1.3.0/go.mod h1:7hK45KPybAkOC6peb+G5yklZfMxEjkZhHbwpqxOKXbg=
 github.com/jonboulle/clockwork v0.1.0 h1:VKV+ZcuP6l3yW9doeqz6ziZGgcynBVQO+obU0+0hcPo=
 github.com/jonboulle/clockwork v0.1.0/go.mod h1:Ii8DK3G1RaLaWxj9trq07+26W01tbo22gdxWY5EU2bo=
@@ -353,8 +369,11 @@ github.com/jtolds/gls v4.20.0+incompatible/go.mod h1:QJZ7F/aHp+rZTRtaJ1ow/lLfFfV
 github.com/julienschmidt/httprouter v1.2.0/go.mod h1:SYymIcj16QtmaHHD7aYtjjsJG7VTCxuUUipMqKk8s4w=
 github.com/karlseguin/ccache v2.0.3+incompatible h1:j68C9tWOROiOLWTS/kCGg9IcJG+ACqn5+0+t8Oh83UU=
 github.com/karlseguin/ccache v2.0.3+incompatible/go.mod h1:CM9tNPzT6EdRh14+jiW8mEF9mkNZuuE51qmgGYUB93w=
+github.com/karlseguin/ccache/v2 v2.0.7 h1:y5Pfi4eiyYCOD6LS/Kj+o6Nb4M5Ngpw9qFQs+v44ZYM=
+github.com/karlseguin/ccache/v2 v2.0.7/go.mod h1:2BDThcfQMf/c0jnZowt16eW405XIqZPavt+HoYEtcxQ=
 github.com/karlseguin/expect v1.0.1 h1:z4wy4npwwHSWKjGWH85WNJO42VQhovxTCZDSzhjo8hY=
 github.com/karlseguin/expect v1.0.1/go.mod h1:zNBxMY8P21owkeogJELCLeHIt+voOSduHYTFUbwRAV8=
+github.com/karlseguin/expect v1.0.2-0.20190806010014-778a5f0c6003/go.mod h1:zNBxMY8P21owkeogJELCLeHIt+voOSduHYTFUbwRAV8=
 github.com/karrick/godirwalk v1.8.0/go.mod h1:H5KPZjojv4lE+QYImBI8xVtrBRgYrIVsaRPx4tDPEn4=
 github.com/karrick/godirwalk v1.10.3/go.mod h1:RoGL9dQei4vP9ilrpETWE8CLOZ1kiN0LhBygSwrAsHA=
 github.com/kisielk/errcheck v1.1.0/go.mod h1:EZBBE59ingxPouuu3KfxchcWSUPOHkagtvWXihfKN4Q=
@@ -453,6 +472,8 @@ github.com/nats-io/nkeys v0.1.3 h1:6JrEfig+HzTH85yxzhSVbjHRJv9cn0p6n3IngIcM5/k=
 github.com/nats-io/nkeys v0.1.3/go.mod h1:xpnFELMwJABBLVhffcfd1MZx6VsNRFpEugbxziKVo7w=
 github.com/nats-io/nuid v1.0.1 h1:5iA8DT8V7q8WK2EScv2padNa/rTESc1KdnPw4TC2paw=
 github.com/nats-io/nuid v1.0.1/go.mod h1:19wcPz3Ph3q0Jbyiqsd0kePYG7A95tJPxeL+1OSON2c=
+github.com/nxadm/tail v1.4.4 h1:DQuhQpB1tVlglWS2hLQ5OV6B5r8aGxSrPc5Qo6uTN78=
+github.com/nxadm/tail v1.4.4/go.mod h1:kenIhsEOeOJmVchQTgglprH7qJGnHDVpk1VPCcaMI8A=
 github.com/oklog/oklog v0.3.2/go.mod h1:FCV+B7mhrz4o+ueLpx+KqkyXRGMWOYEvfiXtdGtbWGs=
 github.com/oklog/run v1.0.0/go.mod h1:dlhp/R75TPv97u0XWUtDeV/lRKWPKSdTuV0TZvrmrQA=
 github.com/oklog/ulid v1.3.1/go.mod h1:CirwcVhetQ6Lv90oh/F+FBtV6XMibvdAFo93nm5qn4U=
@@ -462,12 +483,15 @@ github.com/olivere/elastic/v7 v7.0.19/go.mod h1:4Jqt5xvjqpjCqgnTcHwl3j8TLs8mvoOK
 github.com/onsi/ginkgo v1.6.0/go.mod h1:lLunBs/Ym6LB5Z9jYTR76FiuTmxDTDusOGeTQH+WWjE=
 github.com/onsi/ginkgo v1.7.0 h1:WSHQ+IS43OoUrWtD1/bbclrwK8TTH5hzp+umCiuxHgs=
 github.com/onsi/ginkgo v1.7.0/go.mod h1:lLunBs/Ym6LB5Z9jYTR76FiuTmxDTDusOGeTQH+WWjE=
-github.com/onsi/ginkgo v1.10.1 h1:q/mM8GF/n0shIN8SaAZ0V+jnLPzen6WIVZdiwrRlMlo=
-github.com/onsi/ginkgo v1.10.1/go.mod h1:lLunBs/Ym6LB5Z9jYTR76FiuTmxDTDusOGeTQH+WWjE=
+github.com/onsi/ginkgo v1.12.1/go.mod h1:zj2OWP4+oCPe1qIXoGWkgMRwljMUYCdkwsT2108oapk=
+github.com/onsi/ginkgo v1.14.2 h1:8mVmC9kjFFmA8H4pKMUhcblgifdkOIXPvbhN1T36q1M=
+github.com/onsi/ginkgo v1.14.2/go.mod h1:iSB4RoI2tjJc9BBv4NKIKWKya62Rps+oPG/Lv9klQyY=
 github.com/onsi/gomega v1.4.3 h1:RE1xgDvH7imwFD45h+u2SgIfERHlS2yNG4DObb5BSKU=
 github.com/onsi/gomega v1.4.3/go.mod h1:ex+gbHU/CVuBBDIJjb2X0qEXbFg53c61hWP/1CpauHY=
-github.com/onsi/gomega v1.7.0 h1:XPnZz8VVBHjVsy1vzJmRwIcSwiUO+JFfrv/xGiigmME=
-github.com/onsi/gomega v1.7.0/go.mod h1:ex+gbHU/CVuBBDIJjb2X0qEXbFg53c61hWP/1CpauHY=
+github.com/onsi/gomega v1.7.1/go.mod h1:XdKZgCCFLUoM/7CFJVPcG8C1xQ1AJ0vpAezJrB7JYyY=
+github.com/onsi/gomega v1.10.1/go.mod h1:iN09h71vgCQne3DLsj+A5owkum+a2tYe+TOCB1ybHNo=
+github.com/onsi/gomega v1.10.4 h1:NiTx7EEvBzu9sFOD1zORteLSt3o8gnlvZZwSE9TnY9U=
+github.com/onsi/gomega v1.10.4/go.mod h1:g/HbgYopi++010VEqkFgJHKC09uJiW9UkXvMUuKHUCQ=
 github.com/op/go-logging v0.0.0-20160315200505-970db520ece7/go.mod h1:HzydrMdWErDVzsI23lYNej1Htcns9BCg93Dk0bBINWk=
 github.com/opentracing-contrib/go-observer v0.0.0-20170622124052-a52f23424492/go.mod h1:Ngi6UdF0k5OKD5t5wlmGhe/EDKPoUM3BXZSSfIuJbis=
 github.com/opentracing/basictracer-go v1.0.0/go.mod h1:QfBfYuafItcjQuMwinw9GhYKwFXS9KnPs5lxoYwgW74=
@@ -539,6 +563,8 @@ github.com/rcrowley/go-metrics v0.0.0-20181016184325-3113b8401b8a h1:9ZKAASQSHhD
 github.com/rcrowley/go-metrics v0.0.0-20181016184325-3113b8401b8a/go.mod h1:bCqnVzQkZxMG4s8nGwiZ5l3QUCyqpo9Y+/ZMZ9VjZe4=
 github.com/rcrowley/go-metrics v0.0.0-20190826022208-cac0b30c2563 h1:dY6ETXrvDG7Sa4vE8ZQG4yqWg6UnOcbqTAahkV813vQ=
 github.com/rcrowley/go-metrics v0.0.0-20190826022208-cac0b30c2563/go.mod h1:bCqnVzQkZxMG4s8nGwiZ5l3QUCyqpo9Y+/ZMZ9VjZe4=
+github.com/remyoudompheng/bigfft v0.0.0-20200410134404-eec4a21b6bb0 h1:OdAsTTz6OkFY5QxjkYwrChwuRruF69c169dPK26NUlk=
+github.com/remyoudompheng/bigfft v0.0.0-20200410134404-eec4a21b6bb0/go.mod h1:qqbHyh8v60DhA7CoWK5oRCqLrMHRGoxYCSS9EjAz6Eo=
 github.com/rogpeppe/fastuuid v0.0.0-20150106093220-6724a57986af/go.mod h1:XWv6SoW27p1b0cqNHllgS5HIMJraePCO15w5zCzIWYg=
 github.com/rogpeppe/go-internal v1.1.0/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
 github.com/rogpeppe/go-internal v1.2.2/go.mod h1:M8bDsm7K2OlrFYOpmOWEs/qY81heoFRclV5y23lUDJ4=
@@ -547,8 +573,14 @@ github.com/russross/blackfriday/v2 v2.0.1/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQD
 github.com/ryanuber/columnize v0.0.0-20160712163229-9b3edd62028f/go.mod h1:sm1tb6uqfes/u+d4ooFouqFdy9/2g9QGwK3SQygK0Ts=
 github.com/samuel/go-zookeeper v0.0.0-20190923202752-2cc03de413da/go.mod h1:gi+0XIa01GRL2eRQVjQkKGqKF3SF9vZR/HnPullcV2E=
 github.com/sean-/seed v0.0.0-20170313163322-e2103e2c3529/go.mod h1:DxrIzT+xaE7yg65j358z/aeFdxmN0P9QXhEzd20vsDc=
-github.com/seaweedfs/fuse v1.0.7 h1:tESMXhI3gXzN+dlWsCUrkIZDiWA4dZX18rQMoqmvazw=
-github.com/seaweedfs/fuse v1.0.7/go.mod h1:W7ubwr1l7KQsMeUpxFFOFOSxUL/ucTRMAlVYs4xdfQ8=
+github.com/seaweedfs/fuse v1.0.8 h1:HBPJTC77OlxwSd2JiTwvLPn8bWTElqQp3xs9vf3C15s=
+github.com/seaweedfs/fuse v1.0.8/go.mod h1:W7ubwr1l7KQsMeUpxFFOFOSxUL/ucTRMAlVYs4xdfQ8=
+github.com/seaweedfs/fuse v1.0.9 h1:3JZoGsW7cmmrd/U5KYcIGR2+EqyBvCYCoCpEdZAz/Dc=
+github.com/seaweedfs/fuse v1.0.9/go.mod h1:+PP6WlkrRUG6KPE+Th2EX5To/PjHaFsvqg/UgQ39aj8=
+github.com/seaweedfs/fuse v1.1.0 h1:cL1qPHFNtFv0UuJTLjKKgWDzfJ4iZzTa4Y7ipc2acGw=
+github.com/seaweedfs/fuse v1.1.0/go.mod h1:+PP6WlkrRUG6KPE+Th2EX5To/PjHaFsvqg/UgQ39aj8=
+github.com/seaweedfs/fuse v1.1.1 h1:WD51YFJcBViOx8I89jeqPD+vAKl4EowzBy9GUw0plb0=
+github.com/seaweedfs/fuse v1.1.1/go.mod h1:+PP6WlkrRUG6KPE+Th2EX5To/PjHaFsvqg/UgQ39aj8=
 github.com/seaweedfs/goexif v1.0.2 h1:p+rTXYdQ2mgxd+1JaTrQ9N8DvYuw9UH9xgYmJ+Bb29E=
 github.com/seaweedfs/goexif v1.0.2/go.mod h1:MrKs5LK0HXdffrdCZrW3OIMegL2xXpC6ThLyXMyjdrk=
 github.com/secsy/goftp v0.0.0-20190720192957-f31499d7c79a h1:C6IhVTxNkhlb0tlCB6JfHOUv1f0xHPK7V8X4HlJZEJw=
@@ -560,6 +592,8 @@ github.com/sirupsen/logrus v1.4.0/go.mod h1:LxeOpSwHxABJmUn/MG1IvRgCAasNZTLOkJPx
 github.com/sirupsen/logrus v1.4.1/go.mod h1:ni0Sbl8bgC9z8RoU9G6nDWqqs/fq4eDPysMBDgk/93Q=
 github.com/sirupsen/logrus v1.4.2 h1:SPIRibHv4MatM3XXNO2BJeFLZwZ2LvZgfQ5+UNI2im4=
 github.com/sirupsen/logrus v1.4.2/go.mod h1:tLMulIdttU9McNUspp0xgXVQah82FyeX6MwdIuYE2rE=
+github.com/sirupsen/logrus v1.5.0 h1:1N5EYkVAPEywqZRJd7cwnRtCb6xJx7NH3T3WUTF980Q=
+github.com/sirupsen/logrus v1.5.0/go.mod h1:+F7Ogzej0PZc/94MaYx/nvG9jOFMD2osvC3s+Squfpo=
 github.com/skip2/go-qrcode v0.0.0-20200617195104-da1b6568686e h1:MRM5ITcdelLK2j1vwZ3Je0FKVCfqOLp5zO6trqMLYs0=
 github.com/skip2/go-qrcode v0.0.0-20200617195104-da1b6568686e/go.mod h1:XV66xRDqSt+GTGFMVlhk3ULuV0y9ZmzeVGR4mloJI3M=
 github.com/smartystreets/assertions v0.0.0-20180927180507-b2de0cb4f26d/go.mod h1:OnSkiWE9lh6wB0YB77sQom3nweQdgAjqCqsofrRNTgc=
@@ -610,6 +644,8 @@ github.com/stretchr/testify v1.6.1 h1:hDPOHmpOpP40lSULcqw7IrRb/u7w6RpDC9399XyoNd
 github.com/stretchr/testify v1.6.1/go.mod h1:6Fq8oRcR53rry900zMqJjRRixrwX3KX962/h/Wwjteg=
 github.com/syndtr/goleveldb v1.0.0 h1:fBdIW9lB4Iz0n9khmH8w27SJ3QEJ7+IgjPEwGSZiFdE=
 github.com/syndtr/goleveldb v1.0.0/go.mod h1:ZVVdQEZoIme9iO1Ch2Jdy24qqXrMMOU6lpPAyBWyWuQ=
+github.com/tecbot/gorocksdb v0.0.0-20191217155057-f0fad39f321c h1:g+WoO5jjkqGAzHWCjJB1zZfXPIAaDpzXIEJ0eS6B5Ok=
+github.com/tecbot/gorocksdb v0.0.0-20191217155057-f0fad39f321c/go.mod h1:ahpPrc7HpcfEWDQRZEmnXMzHY03mLDYMCxeDzy46i+8=
 github.com/tidwall/gjson v1.3.2 h1:+7p3qQFaH3fOMXAJSrdZwGKcOO/lYdGS0HqGhPqDdTI=
 github.com/tidwall/gjson v1.3.2/go.mod h1:P256ACg0Mn+j1RXIDXoss50DeIABTYK1PULOJHhxOls=
 github.com/tidwall/match v1.0.1 h1:PnKP62LPNxHKTwvHHZZzdOAOCtsJTjo6dZLCwpKm5xc=
@@ -619,6 +655,8 @@ github.com/tidwall/pretty v1.0.0/go.mod h1:XNkn88O1ChpSDQmQeStsy+sBenx6DDtFZJxhV
 github.com/tmc/grpc-websocket-proxy v0.0.0-20170815181823-89b8d40f7ca8/go.mod h1:ncp9v5uamzpCO7NfCPTXjqaC+bZgJeR0sMTm6dMHP7U=
 github.com/tmc/grpc-websocket-proxy v0.0.0-20190109142713-0ad062ec5ee5 h1:LnC5Kc/wtumK+WB441p7ynQJzVuNRJiqddSIE3IlSEQ=
 github.com/tmc/grpc-websocket-proxy v0.0.0-20190109142713-0ad062ec5ee5/go.mod h1:ncp9v5uamzpCO7NfCPTXjqaC+bZgJeR0sMTm6dMHP7U=
+github.com/tsuna/gohbase v0.0.0-20201125011725-348991136365 h1:6iRwZdrFUzbcVYZwa8dXTIILGIxmmhjyUPJEcwzPGaU=
+github.com/tsuna/gohbase v0.0.0-20201125011725-348991136365/go.mod h1:zj0GJHGvyf1ed3Jm/Tb4830c/ZKDq+YoLsCt2rGQuT0=
 github.com/ugorji/go v1.1.4 h1:j4s+tAvLfL3bZyefP2SEWmhBzmuIlH/eqNuPdFPgngw=
 github.com/ugorji/go v1.1.4/go.mod h1:uQMGLiO92mf5W77hV/PUCpI3pbzQx3CRekS0kk+RGrc=
 github.com/urfave/cli v1.20.0/go.mod h1:70zkFmudgCuE/ngEzBv17Jvp/497gISqfk5gWijbERA=
@@ -662,6 +700,8 @@ go.opencensus.io v0.22.0/go.mod h1:+kGneAE2xo2IficOXnaByMWTGM9T73dGwxeWcUqIpI8=
 go.opencensus.io v0.22.2/go.mod h1:yxeiOL68Rb0Xd1ddK5vPZ/oVn4vY4Ynel7k9FzqtOIw=
 go.opencensus.io v0.22.4 h1:LYy1Hy3MJdrCdMwwzxA/dRok4ejH+RwNGbuoD9fCjto=
 go.opencensus.io v0.22.4/go.mod h1:yxeiOL68Rb0Xd1ddK5vPZ/oVn4vY4Ynel7k9FzqtOIw=
+go.opentelemetry.io/otel v0.15.0 h1:CZFy2lPhxd4HlhZnYK8gRyDotksO3Ip9rBweY1vVYJw=
+go.opentelemetry.io/otel v0.15.0/go.mod h1:e4GKElweB8W2gWUqbghw0B8t5MCTccc9212eNHnOHwA=
 go.uber.org/atomic v1.4.0 h1:cxzIVoETapQEqDhQu3QfnvXAV4AlzcvUCxkVUFw3+EU=
 go.uber.org/atomic v1.4.0/go.mod h1:gD2HeocX3+yG+ygLZcrzQJaqmWj9AIm7n08wl/qW/PE=
 go.uber.org/atomic v1.5.0/go.mod h1:sABNBOSYdrvTF6hTgEIbc7YasKWGhgEQZyfxyTvoXHQ=
@@ -701,6 +741,8 @@ golang.org/x/crypto v0.0.0-20191002192127-34f69633bfdc h1:c0o/qxkaO2LF5t6fQrT4b5
 golang.org/x/crypto v0.0.0-20191002192127-34f69633bfdc/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
 golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550 h1:ObdrDkeb4kJdCP557AjRjq69pTHfNouLtWZG7j9rPN8=
 golang.org/x/crypto v0.0.0-20191011191535-87dc89f01550/go.mod h1:yigFU9vqHzYiE8UmvKecakEJjdnWj3jj499lnFckfCI=
+golang.org/x/crypto v0.0.0-20200622213623-75b288015ac9 h1:psW17arqaxU48Z5kZ0CQnkZWQJsqcURM6tKiBApRjXI=
+golang.org/x/crypto v0.0.0-20200622213623-75b288015ac9/go.mod h1:LzIPMQfyMNhhGPhUkYOs5KpL4U8rLKemX1yGLhDgUto=
 golang.org/x/exp v0.0.0-20190121172915-509febef88a4/go.mod h1:CJ0aWSM057203Lf6IL+f9T1iT9GByDxfZKAQTCR3kQA=
 golang.org/x/exp v0.0.0-20190510132918-efd6b22b2522/go.mod h1:ZjyILWgesfNpC6sMxTJOJm9Kp84zZh5NQWvqDGG3Qr8=
 golang.org/x/image v0.0.0-20190227222117-0694c2d4d067 h1:KYGJGHOQy8oSi1fDlSpcZF0+juKwk/hEMv5SiwHogR0=
@@ -740,6 +782,9 @@ golang.org/x/net v0.0.0-20190620200207-3b0461eec859/go.mod h1:z5CRVTTTmAJ677TzLL
 golang.org/x/net v0.0.0-20190813141303-74dc4d7220e7/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
 golang.org/x/net v0.0.0-20200202094626-16171245cfb2 h1:CCH4IOTTfewWjGOlSp+zGcjutRKlBEZQ6wTn8ozI/nI=
 golang.org/x/net v0.0.0-20200202094626-16171245cfb2/go.mod h1:z5CRVTTTmAJ677TzLLGU+0bjPO0LkuOLi4/5GtJWs/s=
+golang.org/x/net v0.0.0-20200520004742-59133d7f0dd7/go.mod h1:qpuaurCH72eLCgpAm/N6yyVIVM9cpaDIP3A8BGJEC5A=
+golang.org/x/net v0.0.0-20201202161906-c7110b5ffcbb h1:eBmm0M9fYhWpKZLjQUUKka/LtIxf46G4fxeEz5KJr9U=
+golang.org/x/net v0.0.0-20201202161906-c7110b5ffcbb/go.mod h1:sp8m0HH+o8qH0wwXwYZr8TS3Oi6o0r6Gce1SSxlDquU=
 golang.org/x/oauth2 v0.0.0-20180821212333-d2e6202438be/go.mod h1:N/0e6XlmueqKjAGxoOufVs8QHGRruUQn6yWY3a++T0U=
 golang.org/x/oauth2 v0.0.0-20190226205417-e64efc72b421/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
 golang.org/x/oauth2 v0.0.0-20190402181905-9f3314589c9a/go.mod h1:gOpvHmFTYa4IltrdGE7lF6nIHvwfUNPOp7c8zoXwtLw=
@@ -778,15 +823,25 @@ golang.org/x/sys v0.0.0-20190620070143-6f217b454f45/go.mod h1:h1NjWce9XRLGQEsW7w
 golang.org/x/sys v0.0.0-20190624142023-c5567b49c5d0 h1:HyfiK1WMnHj5FXFXatD+Qs1A/xC2Run6RzeW1SyHxpc=
 golang.org/x/sys v0.0.0-20190624142023-c5567b49c5d0/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20190726091711-fc99dfbffb4e/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20190904154756-749cb33beabd/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20191005200804-aed5e4c7ecf9/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20191120155948-bd437916bb0e/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20191220142924-d4481acd189f/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20200202164722-d101bd2416d5 h1:LfCXLvNmTYH9kEmVgqbnsWfruoXZIrh4YBgqVHtDvw0=
 golang.org/x/sys v0.0.0-20200202164722-d101bd2416d5/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20200323222414-85ca7c5b95cd/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20200420163511-1957bb5e6d1f/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20200519105757-fe76b779f299/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/sys v0.0.0-20200930185726-fdedc70b468f/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
 golang.org/x/sys v0.0.0-20201022201747-fb209a7c41cd h1:WgqgiQvkiZWz7XLhphjt2GI2GcGCTIZs9jqXMWmH+oc=
 golang.org/x/sys v0.0.0-20201022201747-fb209a7c41cd/go.mod h1:h1NjWce9XRLGQEsW7wpKNCjG9DtNlClVuFLEZdDNbEs=
+golang.org/x/text v0.0.0-20170915032832-14c0d48ead0c/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.0/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.1-0.20180807135948-17ff2d5776d2/go.mod h1:NqM8EUOU14njkJ3fqMW+pc6Ldnwhi/IjpwHt7yyuwOQ=
 golang.org/x/text v0.3.2 h1:tW2bmiBqwgJj/UpqtC8EpXEZVYOwU0yG4iWbprSVAcs=
 golang.org/x/text v0.3.2/go.mod h1:bEr9sfX3Q8Zfm5fL9x+3itogRgK3+ptLWKqgva+5dAk=
+golang.org/x/text v0.3.3 h1:cokOdA+Jmi5PJGXLlLllQSgYigAEfHXJAERHVMaCc2k=
+golang.org/x/text v0.3.3/go.mod h1:5Zoc/QRtKVWzQhOtBMvqHzDpF6irO9z98xDceosuGiQ=
 golang.org/x/time v0.0.0-20180412165947-fbb02b2291d2/go.mod h1:tRJNPiyCQ0inRvYxbN9jk5I+vvW/OXSQhTDSoE431IQ=
 golang.org/x/time v0.0.0-20181108054448-85acf8d2951c/go.mod h1:tRJNPiyCQ0inRvYxbN9jk5I+vvW/OXSQhTDSoE431IQ=
 golang.org/x/time v0.0.0-20190308202827-9d24e82272b4 h1:SvFZT6jyqRaOeXpc5h/JSfZenJ2O330aBsf7JfSUXmQ=
@@ -919,6 +974,11 @@ gopkg.in/yaml.v2 v2.0.0-20170812160011-eb3733d160e7/go.mod h1:JAlM8MvJe8wmxCU4Bl
 gopkg.in/yaml.v2 v2.2.1/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
 gopkg.in/yaml.v2 v2.2.2 h1:ZCJp+EgiOT7lHqUV2J862kp8Qj64Jo6az82+3Td9dZw=
 gopkg.in/yaml.v2 v2.2.2/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
+gopkg.in/yaml.v2 v2.2.4/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
+gopkg.in/yaml.v2 v2.2.8 h1:obN1ZagJSUGI0Ek/LBmuj4SNLPfIny3KsKFopxRdj10=
+gopkg.in/yaml.v2 v2.2.8/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
+gopkg.in/yaml.v2 v2.3.0 h1:clyUAQHOM3G0M3f5vQj7LuJrETvjVot3Z5el9nffUtU=
+gopkg.in/yaml.v2 v2.3.0/go.mod h1:hI93XBmqTisBFMUTm0b8Fm+jr3Dg1NNxqwp+5A1VGuI=
 gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c h1:dUUwHk2QECo/6vqA44rthZ8ie2QXMNeKRTHCNY2nXvo=
 gopkg.in/yaml.v3 v3.0.0-20200313102051-9f266ea9e77c/go.mod h1:K4uyk7z7BCEPqu6E+C64Yfv1cQ7kz7rIZviUmN+EgEM=
 honnef.co/go/tools v0.0.0-20180728063816-88497007e858/go.mod h1:rf3lG4BRIbNafJWhAfAdb/ePZxsR/4RtNHQocxwk9r4=
@@ -928,8 +988,16 @@ honnef.co/go/tools v0.0.0-20190418001031-e561f6794a2a/go.mod h1:rf3lG4BRIbNafJWh
 honnef.co/go/tools v0.0.0-20190523083050-ea95bdfd59fc/go.mod h1:rf3lG4BRIbNafJWhAfAdb/ePZxsR/4RtNHQocxwk9r4=
 honnef.co/go/tools v0.0.1-2019.2.3 h1:3JgtbtFHMiCmsznwGVTUWbgGov+pVqnlf1dEJTNAXeM=
 honnef.co/go/tools v0.0.1-2019.2.3/go.mod h1:a3bituU0lyd329TUQxRnasdCoJDkEUEAqEt0JzvZhAg=
+modernc.org/b v1.0.0 h1:vpvqeyp17ddcQWF29Czawql4lDdABCDRbXRAS4+aF2o=
+modernc.org/b v1.0.0/go.mod h1:uZWcZfRj1BpYzfN9JTerzlNUnnPsV9O2ZA8JsRcubNg=
+modernc.org/mathutil v1.1.1 h1:FeylZSVX8S+58VsyJlkEj2bcpdytmp9MmDKZkKx8OIE=
+modernc.org/mathutil v1.1.1/go.mod h1:mZW8CKdRPY1v87qxC/wUdX5O1qDzXMP5TH3wjfpga6E=
+modernc.org/strutil v1.1.0 h1:+1/yCzZxY2pZwwrsbH+4T7BQMoLQ9QiBshRC9eicYsc=
+modernc.org/strutil v1.1.0/go.mod h1:lstksw84oURvj9y3tn8lGvRxyRC1S2+g5uuIzNfIOBs=
 pack.ag/amqp v0.11.2/go.mod h1:4/cbmt4EJXSKlG6LCfWHoqmN0uFdy5i/+YFz+fTfhV4=
 rsc.io/binaryregexp v0.2.0/go.mod h1:qTv7/COck+e2FymRvadv62gMdZztPaShugOCi3I+8D8=
+rsc.io/quote/v3 v3.1.0/go.mod h1:yEA65RcK8LyAZtP9Kv3t0HmxON59tX3rD+tICJqUlj0=
+rsc.io/sampler v1.3.0/go.mod h1:T1hPZKmBbMNahiBKFy5HrXp6adAjACjK9JXDnKaTXpA=
 sigs.k8s.io/yaml v1.1.0 h1:4A07+ZFc2wgJwo8YNlQpr1rVlgUDlxXHhPJciaPY5gs=
 sigs.k8s.io/yaml v1.1.0/go.mod h1:UJmg0vDUVViEyp3mgSv9WPwZCDxu4rQW1olrI1uml+o=
 sourcegraph.com/sourcegraph/appdash v0.0.0-20190731080439-ebfcffb1b5c0/go.mod h1:hI742Nqp5OhwiqlzhgfbWU4mW4yO10fP+LoT9WOswdU=
@@ -1,4 +1,5 @@
 apiVersion: v1
 description: SeaweedFS
 name: seaweedfs
-version: 2.12
+appVersion: "2.23"
+version: 2.23
@@ -52,12 +52,12 @@ Inject extra environment vars in the format key:value, if populated
 {{- $registryName := default .Values.image.registry .Values.global.localRegistry | toString -}}
 {{- $repositoryName := .Values.image.repository | toString -}}
 {{- $name := .Values.global.imageName | toString -}}
-{{- $tag := .Values.global.imageTag | toString -}}
+{{- $tag := .Chart.AppVersion | toString -}}
 {{- printf "%s%s%s:%s" $registryName $repositoryName $name $tag -}}
 {{- end -}}
 {{- end -}}

-{{/* Return the proper postgresqlSchema image */}}
+{{/* Return the proper dbSchema image */}}
 {{- define "filer.dbSchema.image" -}}
 {{- if .Values.filer.dbSchema.imageOverride -}}
 {{- $imageOverride := .Values.filer.dbSchema.imageOverride -}}
@@ -80,7 +80,7 @@ Inject extra environment vars in the format key:value, if populated
 {{- $registryName := default .Values.image.registry .Values.global.localRegistry | toString -}}
 {{- $repositoryName := .Values.image.repository | toString -}}
 {{- $name := .Values.global.imageName | toString -}}
-{{- $tag := .Values.global.imageTag | toString -}}
+{{- $tag := .Chart.AppVersion | toString -}}
 {{- printf "%s%s%s:%s" $registryName $repositoryName $name $tag -}}
 {{- end -}}
 {{- end -}}
@@ -94,7 +94,7 @@ Inject extra environment vars in the format key:value, if populated
 {{- $registryName := default .Values.image.registry .Values.global.localRegistry | toString -}}
 {{- $repositoryName := .Values.image.repository | toString -}}
 {{- $name := .Values.global.imageName | toString -}}
-{{- $tag := .Values.global.imageTag | toString -}}
+{{- $tag := .Chart.AppVersion | toString -}}
 {{- printf "%s%s%s:%s" $registryName $repositoryName $name $tag -}}
 {{- end -}}
 {{- end -}}
@@ -108,7 +108,7 @@ Inject extra environment vars in the format key:value, if populated
 {{- $registryName := default .Values.image.registry .Values.global.localRegistry | toString -}}
 {{- $repositoryName := .Values.image.repository | toString -}}
 {{- $name := .Values.global.imageName | toString -}}
-{{- $tag := .Values.global.imageTag | toString -}}
+{{- $tag := .Chart.AppVersion | toString -}}
 {{- printf "%s%s%s:%s" $registryName $repositoryName $name $tag -}}
 {{- end -}}
 {{- end -}}
@@ -122,7 +122,7 @@ Inject extra environment vars in the format key:value, if populated
 {{- $registryName := default .Values.image.registry .Values.global.localRegistry | toString -}}
 {{- $repositoryName := .Values.image.repository | toString -}}
 {{- $name := .Values.global.imageName | toString -}}
-{{- $tag := .Values.global.imageTag | toString -}}
+{{- $tag := .Chart.AppVersion | toString -}}
 {{- printf "%s%s%s:%s" $registryName $repositoryName $name $tag -}}
 {{- end -}}
-{{- end -}}
+{{- end -}}
@@ -6,6 +6,7 @@ metadata:
  name: {{ include "seaweedfs.fullname" . }}-cronjob
 spec:
  schedule: "{{ .Values.cronjob.schedule }}"
+  startingDeadlineSeconds: 200
  concurrencyPolicy: Forbid
  failedJobsHistoryLimit: 2
  successfulJobsHistoryLimit: 2
@@ -51,4 +52,4 @@ spec:
                  -filer {{ template "seaweedfs.name" . }}-filer.{{ .Release.Namespace }}.svc:{{ .Values.filer.port }}
                  {{- end }}
 {{- end }}
-{{- end }}
+{{- end }}
@@ -0,0 +1,31 @@
+apiVersion: v1
+kind: Service
+metadata:
+  name: {{ template "seaweedfs.name" . }}-filer-client
+  namespace: {{ .Release.Namespace }}
+  labels:
+    app: {{ template "seaweedfs.name" . }}
+    component: filer
+{{- if .Values.filer.metricsPort }}
+    monitoring: "true"
+{{- end }}
+spec:
+  clusterIP: None
+  ports:
+  - name: "swfs-filer"
+    port: {{ .Values.filer.port }}
+    targetPort: {{ .Values.filer.port }}
+    protocol: TCP
+  - name: "swfs-filer-grpc"
+    port: {{ .Values.filer.grpcPort }}
+    targetPort: {{ .Values.filer.grpcPort }}
+    protocol: TCP
+{{- if .Values.filer.metricsPort }}
+  - name: "metrics"
+    port: {{ .Values.filer.metricsPort }}
+    targetPort: {{ .Values.filer.metricsPort }}
+    protocol: TCP
+{{- end }}
+  selector:
+    app: {{ template "seaweedfs.name" . }}
+    component: filer
@@ -1,6 +1,8 @@
 apiVersion: v1
 kind: Service
 metadata:
+  annotations:
+    service.alpha.kubernetes.io/tolerate-unready-endpoints: "true"
  name: {{ template "seaweedfs.name" . }}-filer
  namespace: {{ .Release.Namespace }}
  labels:
@@ -8,6 +10,7 @@ metadata:
    component: filer
 spec:
  clusterIP: None
+  publishNotReadyAddresses: true
  ports:
  - name: "swfs-filer"
    port: {{ .Values.filer.port }}
@@ -17,12 +20,6 @@ spec:
    port: {{ .Values.filer.grpcPort }}
    targetPort: {{ .Values.filer.grpcPort }}
    protocol: TCP
-{{- if .Values.filer.metricsPort }}
-  - name: "swfs-filer-metrics"
-    port: {{ .Values.filer.metricsPort }}
-    targetPort: {{ .Values.filer.metricsPort }}
-    protocol: TCP
-{{- end }}
  selector:
    app: {{ template "seaweedfs.name" . }}
-    component: filer
+    component: filer
@@ -87,6 +87,12 @@ spec:
              value: {{ $value | quote }}
            {{- end }}
            {{- end }}
+            {{- if .Values.global.extraEnvironmentVars }}
+            {{- range $key, $value := .Values.global.extraEnvironmentVars }}
+            - name: {{ $key }}
+              value: {{ $value | quote }}
+            {{- end }}
+            {{- end }}
          command:
            - "/bin/sh"
            - "-ec"
@@ -128,7 +134,7 @@ spec:
              {{- end }}
              -ip=${POD_IP} \
              {{- if gt (.Values.filer.replicas | int) 1 }}
-              -peers={{ range $index := until (.Values.filer.replicas | int) }}${SEAWEEDFS_FULLNAME}-filer-{{ $index }}.${SEAWEEDFS_FULLNAME}-filer:{{ $.Values.filer.port }}{{ if lt $index (sub ($.Values.filer.replicas | int) 1) }},{{ end }}{{ end }}
+              -peers=$(echo -n "{{ range $index := until (.Values.filer.replicas | int) }}${SEAWEEDFS_FULLNAME}-filer-{{ $index }}.${SEAWEEDFS_FULLNAME}-filer:{{ $.Values.filer.port }}{{ if lt $index (sub ($.Values.filer.replicas | int) 1) }},{{ end }}{{ end }}" | sed "s/$HOSTNAME.${SEAWEEDFS_FULLNAME}-filer:{{ $.Values.filer.port }}//" | sed 's/,$//; 's/^,//'; s/,,/,/;' )  \
              {{- end }}
              -master={{ range $index := until (.Values.master.replicas | int) }}${SEAWEEDFS_FULLNAME}-master-{{ $index }}.${SEAWEEDFS_FULLNAME}-master:{{ $.Values.master.port }}{{ if lt $index (sub ($.Values.master.replicas | int) 1) }},{{ end }}{{ end }}
          {{- if or (.Values.global.enableSecurity) (.Values.filer.extraVolumeMounts) }}    
@@ -172,7 +178,7 @@ spec:
            periodSeconds: 15
            successThreshold: 1
            failureThreshold: 100
-            timeoutSeconds: 3
+            timeoutSeconds: 10
          livenessProbe:
            httpGet:
              path: /
@@ -182,7 +188,7 @@ spec:
            periodSeconds: 30
            successThreshold: 1
            failureThreshold: 5
-            timeoutSeconds: 3
+            timeoutSeconds: 10
          {{- if .Values.filer.resources }}
          resources:
            {{ tpl .Values.filer.resources . | nindent 12 | trim }}
@@ -10,6 +10,7 @@ metadata:
    service.alpha.kubernetes.io/tolerate-unready-endpoints: "true"    
 spec:
  clusterIP: None
+  publishNotReadyAddresses: true
  ports:
  - name: "swfs-master"
    port: {{ .Values.master.port }}
@@ -21,4 +22,4 @@ spec:
    protocol: TCP
  selector:
    app: {{ template "seaweedfs.name" . }}
-    component: master
+    component: master
@@ -76,6 +76,12 @@ spec:
              value: {{ $value | quote }}
            {{- end }}
            {{- end }}
+            {{- if .Values.global.extraEnvironmentVars }}
+            {{- range $key, $value := .Values.global.extraEnvironmentVars }}
+            - name: {{ $key }}
+              value: {{ $value | quote }}
+            {{- end }}
+            {{- end }}
          command:
            - "/bin/sh"
            - "-ec"
@@ -157,7 +163,7 @@ spec:
            periodSeconds: 45
            successThreshold: 2
            failureThreshold: 100
-            timeoutSeconds: 5
+            timeoutSeconds: 10
          livenessProbe:
            httpGet:
              path: /cluster/status
@@ -167,7 +173,7 @@ spec:
            periodSeconds: 30
            successThreshold: 1
            failureThreshold: 4
-            timeoutSeconds: 5
+            timeoutSeconds: 10
          {{- if .Values.master.resources }}
          resources:
            {{ tpl .Values.master.resources . | nindent 12 | trim }}
@@ -59,11 +59,17 @@ spec:
                  fieldPath: metadata.namespace
            - name: SEAWEEDFS_FULLNAME
              value: "{{ template "seaweedfs.name" . }}"
+            {{- if .Values.global.extraEnvironmentVars }}
+            {{- range $key, $value := .Values.global.extraEnvironmentVars }}
+            - name: {{ $key }}
+              value: {{ $value | quote }}
+            {{- end }}
+            {{- end }}
          command:
            - "/bin/sh"
            - "-ec"
            - | 
-              exec /usr/bin/weed \
+              exec /usr/bin/weed -logdir=/logs \
              {{- if .Values.s3.loggingOverrideLevel }}
              -v={{ .Values.s3.loggingOverrideLevel }} \
              {{- else }}
@@ -81,9 +87,13 @@ spec:
              {{- if .Values.s3.domainName }}
              -domainName={{ .Values.s3.domainName }} \
              {{- end }}
-              -filer={{ template "seaweedfs.name" . }}-filer:{{ .Values.filer.port }}
-          {{- if or (.Values.global.enableSecurity) (.Values.s3.extraVolumeMounts) }}
+              {{- if .Values.s3.allowEmptyFolder }}
+              -allowEmptyFolder={{ .Values.s3.allowEmptyFolder }} \
+              {{- end }}
+              -filer={{ template "seaweedfs.name" . }}-filer-client:{{ .Values.filer.port }}
          volumeMounts:
+            - name: logs
+              mountPath: "/logs/"
            {{- if .Values.global.enableSecurity }}
            - name: security-config
              readOnly: true
@@ -106,7 +116,6 @@ spec:
              mountPath: /usr/local/share/ca-certificates/client/
            {{- end }}
            {{ tpl .Values.s3.extraVolumeMounts . | nindent 12 | trim }}
-          {{- end }}
          ports:
            - containerPort: {{ .Values.s3.port }}
              name: swfs-s3
@@ -119,7 +128,7 @@ spec:
            periodSeconds: 15
            successThreshold: 1
            failureThreshold: 100
-            timeoutSeconds: 3
+            timeoutSeconds: 10
          livenessProbe:
            httpGet:
              path: /
@@ -129,12 +138,18 @@ spec:
            periodSeconds: 60
            successThreshold: 1
            failureThreshold: 20
-            timeoutSeconds: 3
+            timeoutSeconds: 10
          {{- if .Values.s3.resources }}
          resources:
            {{ tpl .Values.s3.resources . | nindent 12 | trim }}
          {{- end }}
      volumes:
+        {{- if eq .Values.s3.logs.type "hostPath" }}
+        - name: logs
+          hostPath:
+            path: /storage/logs/seaweedfs/s3
+            type: DirectoryOrCreate
+        {{- end }}
        {{- if .Values.global.enableSecurity }}
        - name: security-config
          configMap:
@@ -64,6 +64,12 @@ spec:
                  fieldPath: status.hostIP
            - name: SEAWEEDFS_FULLNAME
              value: "{{ template "seaweedfs.name" . }}"
+            {{- if .Values.global.extraEnvironmentVars }}
+            {{- range $key, $value := .Values.global.extraEnvironmentVars }}
+            - name: {{ $key }}
+              value: {{ $value | quote }}
+            {{- end }}
+            {{- end }}
          command:
            - "/bin/sh"
            - "-ec"
@@ -80,6 +86,9 @@ spec:
                -metricsPort {{ .Values.volume.metricsPort }} \
                {{- end }}
                -dir={{ .Values.volume.dir }} \
+                {{- if .Values.volume.dir_idx }}
+                -dir.idx={{ .Values.volume.dir_idx }} \
+                {{- end }}
                -max={{ .Values.volume.maxVolumes }} \
                {{- if .Values.volume.rack }}
                -rack={{ .Values.volume.rack }} \
@@ -149,7 +158,7 @@ spec:
            periodSeconds: 90
            successThreshold: 1
            failureThreshold: 100
-            timeoutSeconds: 5
+            timeoutSeconds: 30
          livenessProbe:
            httpGet:
              path: /status
@@ -159,7 +168,7 @@ spec:
            periodSeconds: 90
            successThreshold: 1
            failureThreshold: 4
-            timeoutSeconds: 5
+            timeoutSeconds: 30
          {{- if .Values.volume.resources }}
          resources:
            {{ tpl .Values.volume.resources . | nindent 12 | trim }}
@@ -4,7 +4,7 @@ global:
  registry: ""
  repository: ""
  imageName: chrislusf/seaweedfs
-  imageTag: "2.12"
+  # imageTag: "2.23" - started using {.Chart.appVersion}
  imagePullPolicy: IfNotPresent
  imagePullSecrets: imagepullsecret
  restartPolicy: Always
@@ -21,6 +21,10 @@ global:
  # Y number of replica in other racks in the same data center
  # Z number of replica in other servers in the same rack
  replicationPlacment: "001"
+  extraEnvironmentVars:
+    WEED_CLUSTER_DEFAULT: "sw"
+    WEED_CLUSTER_SW_MASTER: "seaweedfs-master:9333"
+    WEED_CLUSTER_SW_FILER: "seaweedfs-filer-client:8888"

 image:
  registry: ""
@@ -140,6 +144,8 @@ volume:

  # Directories to store data files. dir[,dir]... (default "/tmp")
  dir: "/data"
+  # Directories to store index files. dir[,dir]... (default "/tmp")
+  dir_idx: null

  # Maximum numbers of volumes, count[,count]...
  # If set to zero on non-windows OS, the limit will be auto configured. (default "7")
@@ -292,8 +298,10 @@ filer:
    WEED_MYSQL_HOSTNAME: "mysql-db-host"
    WEED_MYSQL_PORT: "3306"
    WEED_MYSQL_DATABASE: "sw_database"
-    WEED_MYSQL_CONNECTION_MAX_IDLE: "10"
-    WEED_MYSQL_CONNECTION_MAX_OPEN: "150"
+    WEED_MYSQL_CONNECTION_MAX_IDLE: "5"
+    WEED_MYSQL_CONNECTION_MAX_OPEN: "75"
+    # "refresh" connection every 10 minutes, eliminating mysql closing "old" connections
+    WEED_MYSQL_CONNECTION_MAX_LIFETIME_SECONDS: "600"
    # enable usage of memsql as filer backend
    WEED_MYSQL_INTERPOLATEPARAMS: "true"
    WEED_LEVELDB2_ENABLED: "false"
@@ -313,6 +321,8 @@ s3:
  port: 8333
  metricsPort: 9327
  loggingOverrideLevel: null
+  #allow empty folders
+  allowEmptyFolder: true

  # Suffix of the host name, {bucket}.{domainName}
  domainName: ""
@@ -343,6 +353,11 @@ s3:
  # ref: https://kubernetes.io/docs/concepts/configuration/pod-priority-preemption/
  priorityClassName: ""

+  logs:
+    type: "hostPath"
+    size: ""
+    storageClass: ""
+
 cronjob:
  enabled: false
  schedule: "*/7 * * * *"
@@ -5,7 +5,7 @@

    <groupId>com.github.chrislusf</groupId>
    <artifactId>seaweedfs-client</artifactId>
-    <version>1.5.6</version>
+    <version>1.5.7</version>

    <parent>
        <groupId>org.sonatype.oss</groupId>
@@ -5,7 +5,7 @@

    <groupId>com.github.chrislusf</groupId>
    <artifactId>seaweedfs-client</artifactId>
-    <version>1.5.6</version>
+    <version>1.5.7</version>

    <parent>
        <groupId>org.sonatype.oss</groupId>
@@ -5,7 +5,7 @@

    <groupId>com.github.chrislusf</groupId>
    <artifactId>seaweedfs-client</artifactId>
-    <version>1.5.6</version>
+    <version>1.5.7</version>

    <parent>
        <groupId>org.sonatype.oss</groupId>
@@ -68,6 +68,11 @@
            <version>4.13.1</version>
            <scope>test</scope>
        </dependency>
+        <dependency>
+            <groupId>javax.annotation</groupId>
+            <artifactId>javax.annotation-api</artifactId>
+            <version>1.3.2</version>
+        </dependency>
    </dependencies>

    <build>
@@ -1,26 +1,22 @@
-package seaweed.hdfs;
+package seaweedfs.client;

 // based on org.apache.hadoop.fs.azurebfs.services.AbfsInputStream

-import org.apache.hadoop.fs.FSExceptionMessages;
-import org.apache.hadoop.fs.FSInputStream;
-import org.apache.hadoop.fs.FileSystem.Statistics;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
-import seaweedfs.client.FilerGrpcClient;
-import seaweedfs.client.FilerProto;
-import seaweedfs.client.SeaweedRead;

 import java.io.EOFException;
 import java.io.IOException;
+import java.io.InputStream;
+import java.nio.ByteBuffer;
 import java.util.List;

-public class SeaweedInputStream extends FSInputStream {
+public class SeaweedInputStream extends InputStream {

    private static final Logger LOG = LoggerFactory.getLogger(SeaweedInputStream.class);
+    private static final IOException EXCEPTION_STREAM_IS_CLOSED = new IOException("Stream is closed!");

    private final FilerGrpcClient filerGrpcClient;
-    private final Statistics statistics;
    private final String path;
    private final FilerProto.Entry entry;
    private final List<SeaweedRead.VisibleInterval> visibleIntervalList;
@@ -32,11 +28,9 @@ public class SeaweedInputStream extends FSInputStream {

    public SeaweedInputStream(
            final FilerGrpcClient filerGrpcClient,
-            final Statistics statistics,
            final String path,
            final FilerProto.Entry entry) throws IOException {
        this.filerGrpcClient = filerGrpcClient;
-        this.statistics = statistics;
        this.path = path;
        this.entry = entry;
        this.contentLength = SeaweedRead.fileSize(entry);
@@ -63,14 +57,8 @@ public class SeaweedInputStream extends FSInputStream {
    }

    @Override
-    public synchronized int read(final byte[] b, final int off, final int len) throws IOException {
+    public int read(final byte[] b, final int off, final int len) throws IOException {

-        if (position < 0) {
-            throw new IllegalArgumentException("attempting to read from negative offset");
-        }
-        if (position >= contentLength) {
-            return -1;  // Hadoop prefers -1 to EOFException
-        }
        if (b == null) {
            throw new IllegalArgumentException("null byte array passed in to read() method");
        }
@@ -84,53 +72,62 @@ public class SeaweedInputStream extends FSInputStream {
            throw new IllegalArgumentException("requested read length is more than will fit after requested offset in buffer");
        }

-        long bytesRead = SeaweedRead.read(this.filerGrpcClient, this.visibleIntervalList, this.position, b, off, len, SeaweedRead.fileSize(entry));
+        ByteBuffer buf = ByteBuffer.wrap(b, off, len);
+        return read(buf);
+
+    }
+
+    // implement ByteBufferReadable
+    public synchronized int read(ByteBuffer buf) throws IOException {
+
+        if (position < 0) {
+            throw new IllegalArgumentException("attempting to read from negative offset");
+        }
+        if (position >= contentLength) {
+            return -1;  // Hadoop prefers -1 to EOFException
+        }
+
+        long bytesRead = 0;
+        int len = buf.remaining();
+        int start = (int) this.position;
+        if (start+len <= entry.getContent().size()) {
+            entry.getContent().substring(start, start+len).copyTo(buf);
+        } else {
+            bytesRead = SeaweedRead.read(this.filerGrpcClient, this.visibleIntervalList, this.position, buf, SeaweedRead.fileSize(entry));
+        }
+
        if (bytesRead > Integer.MAX_VALUE) {
            throw new IOException("Unexpected Content-Length");
        }

        if (bytesRead > 0) {
            this.position += bytesRead;
-            if (statistics != null) {
-                statistics.incrementBytesRead(bytesRead);
-            }
        }

        return (int) bytesRead;
-
    }

-    /**
-     * Seek to given position in stream.
-     *
-     * @param n position to seek to
-     * @throws IOException  if there is an error
-     * @throws EOFException if attempting to seek past end of file
-     */
-    @Override
    public synchronized void seek(long n) throws IOException {
        if (closed) {
-            throw new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
+            throw EXCEPTION_STREAM_IS_CLOSED;
        }
        if (n < 0) {
-            throw new EOFException(FSExceptionMessages.NEGATIVE_SEEK);
+            throw new EOFException("Cannot seek to a negative offset");
        }
        if (n > contentLength) {
-            throw new EOFException(FSExceptionMessages.CANNOT_SEEK_PAST_EOF);
+            throw new EOFException("Attempted to seek or read past the end of the file");
        }
-
        this.position = n;
-
    }

    @Override
    public synchronized long skip(long n) throws IOException {
        if (closed) {
-            throw new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
+            throw EXCEPTION_STREAM_IS_CLOSED;
        }
        if (this.position == contentLength) {
            if (n > 0) {
-                throw new EOFException(FSExceptionMessages.CANNOT_SEEK_PAST_EOF);
+                throw new EOFException("Attempted to seek or read past the end of the file");
            }
        }
        long newPos = this.position + n;
@@ -158,10 +155,9 @@ public class SeaweedInputStream extends FSInputStream {
    @Override
    public synchronized int available() throws IOException {
        if (closed) {
-            throw new IOException(
-                    FSExceptionMessages.STREAM_IS_CLOSED);
+            throw EXCEPTION_STREAM_IS_CLOSED;
        }
-        final long remaining = this.contentLength - this.getPos();
+        final long remaining = this.contentLength - this.position;
        return remaining <= Integer.MAX_VALUE
                ? (int) remaining : Integer.MAX_VALUE;
    }
@@ -176,65 +172,21 @@ public class SeaweedInputStream extends FSInputStream {
     */
    public long length() throws IOException {
        if (closed) {
-            throw new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
+            throw EXCEPTION_STREAM_IS_CLOSED;
        }
        return contentLength;
    }

-    /**
-     * Return the current offset from the start of the file
-     *
-     * @throws IOException throws {@link IOException} if there is an error
-     */
-    @Override
    public synchronized long getPos() throws IOException {
        if (closed) {
-            throw new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
+            throw EXCEPTION_STREAM_IS_CLOSED;
        }
        return position;
    }

-    /**
-     * Seeks a different copy of the data.  Returns true if
-     * found a new source, false otherwise.
-     *
-     * @throws IOException throws {@link IOException} if there is an error
-     */
-    @Override
-    public boolean seekToNewSource(long l) throws IOException {
-        return false;
-    }
-
    @Override
    public synchronized void close() throws IOException {
        closed = true;
    }

-    /**
-     * Not supported by this stream. Throws {@link UnsupportedOperationException}
-     *
-     * @param readlimit ignored
-     */
-    @Override
-    public synchronized void mark(int readlimit) {
-        throw new UnsupportedOperationException("mark()/reset() not supported on this stream");
-    }
-
-    /**
-     * Not supported by this stream. Throws {@link UnsupportedOperationException}
-     */
-    @Override
-    public synchronized void reset() throws IOException {
-        throw new UnsupportedOperationException("mark()/reset() not supported on this stream");
-    }
-
-    /**
-     * gets whether mark and reset are supported by {@code ADLFileInputStream}. Always returns false.
-     *
-     * @return always {@code false}
-     */
-    @Override
-    public boolean markSupported() {
-        return false;
-    }
 }
@@ -1,16 +1,9 @@
-package seaweed.hdfs;
+package seaweedfs.client;

 // adapted from org.apache.hadoop.fs.azurebfs.services.AbfsOutputStream

-import com.google.common.base.Preconditions;
-import org.apache.hadoop.fs.FSExceptionMessages;
-import org.apache.hadoop.fs.Path;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
-import seaweedfs.client.ByteBufferPool;
-import seaweedfs.client.FilerGrpcClient;
-import seaweedfs.client.FilerProto;
-import seaweedfs.client.SeaweedWrite;

 import java.io.IOException;
 import java.io.InterruptedIOException;
@@ -18,20 +11,18 @@ import java.io.OutputStream;
 import java.nio.ByteBuffer;
 import java.util.concurrent.*;

-import static seaweed.hdfs.SeaweedFileSystemStore.getParentDirectory;
-
 public class SeaweedOutputStream extends OutputStream {

    private static final Logger LOG = LoggerFactory.getLogger(SeaweedOutputStream.class);

    private final FilerGrpcClient filerGrpcClient;
-    private final Path path;
+    private final String path;
    private final int bufferSize;
    private final int maxConcurrentRequestCount;
    private final ThreadPoolExecutor threadExecutor;
    private final ExecutorCompletionService<Void> completionService;
    private final FilerProto.Entry.Builder entry;
-    private final boolean supportFlush = false; // true;
+    protected final boolean supportFlush = false; // true;
    private final ConcurrentLinkedDeque<WriteOperation> writeOperations;
    private long position;
    private boolean closed;
@@ -42,7 +33,7 @@ public class SeaweedOutputStream extends OutputStream {
    private long outputIndex;
    private String replication = "000";

-    public SeaweedOutputStream(FilerGrpcClient filerGrpcClient, final Path path, FilerProto.Entry.Builder entry,
+    public SeaweedOutputStream(FilerGrpcClient filerGrpcClient, final String path, FilerProto.Entry.Builder entry,
                               final long position, final int bufferSize, final String replication) {
        this.filerGrpcClient = filerGrpcClient;
        this.replication = replication;
@@ -70,6 +61,14 @@ public class SeaweedOutputStream extends OutputStream {

    }

+    public static String getParentDirectory(String path) {
+        if (path.equals("/")) {
+            return path;
+        }
+        int lastSlashIndex = path.lastIndexOf("/");
+        return path.substring(0, lastSlashIndex);
+    }
+
    private synchronized void flushWrittenBytesToServiceInternal(final long offset) throws IOException {
        try {
            SeaweedWrite.writeMeta(filerGrpcClient, getParentDirectory(path), entry);
@@ -89,7 +88,9 @@ public class SeaweedOutputStream extends OutputStream {
            throws IOException {
        maybeThrowLastError();

-        Preconditions.checkArgument(data != null, "null data");
+        if (data == null) {
+            return;
+        }

        if (off < 0 || length < 0 || length > data.length - off) {
            throw new IndexOutOfBoundsException();
@@ -152,7 +153,7 @@ public class SeaweedOutputStream extends OutputStream {
            flushInternal();
            threadExecutor.shutdown();
        } finally {
-            lastError = new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
+            lastError = new IOException("Stream is closed!");
            ByteBufferPool.release(buffer);
            buffer = null;
            outputIndex = 0;
@@ -185,7 +186,7 @@ public class SeaweedOutputStream extends OutputStream {
        }
        final Future<Void> job = completionService.submit(() -> {
            // System.out.println(path + " is going to save [" + (writePosition) + "," + ((writePosition) + bytesLength) + ")");
-            SeaweedWrite.writeData(entry, replication, filerGrpcClient, writePosition, bufferToWrite.array(), bufferToWrite.position(), bufferToWrite.limit(), path.toUri().getPath());
+            SeaweedWrite.writeData(entry, replication, filerGrpcClient, writePosition, bufferToWrite.array(), bufferToWrite.position(), bufferToWrite.limit(), path);
            // System.out.println(path + " saved [" + (writePosition) + "," + ((writePosition) + bytesLength) + ")");
            ByteBufferPool.release(bufferToWrite);
            return null;
@@ -239,13 +240,13 @@ public class SeaweedOutputStream extends OutputStream {
        }
    }

-    private synchronized void flushInternal() throws IOException {
+    protected synchronized void flushInternal() throws IOException {
        maybeThrowLastError();
        writeCurrentBufferToService();
        flushWrittenBytesToService();
    }

-    private synchronized void flushInternalAsync() throws IOException {
+    protected synchronized void flushInternalAsync() throws IOException {
        maybeThrowLastError();
        writeCurrentBufferToService();
        flushWrittenBytesToServiceAsync();
@@ -278,10 +279,6 @@ public class SeaweedOutputStream extends OutputStream {
        private final long length;

        WriteOperation(final Future<Void> task, final long startOffset, final long length) {
-            Preconditions.checkNotNull(task, "task");
-            Preconditions.checkArgument(startOffset >= 0, "startOffset");
-            Preconditions.checkArgument(length >= 0, "length");
-
            this.task = task;
            this.startOffset = startOffset;
            this.length = length;
@@ -12,6 +12,7 @@ import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;

 import java.io.IOException;
+import java.nio.ByteBuffer;
 import java.util.*;

 public class SeaweedRead {
@@ -23,10 +24,9 @@ public class SeaweedRead {

    // returns bytesRead
    public static long read(FilerGrpcClient filerGrpcClient, List<VisibleInterval> visibleIntervals,
-                            final long position, final byte[] buffer, final int bufferOffset,
-                            final int bufferLength, final long fileSize) throws IOException {
+                            final long position, final ByteBuffer buf, final long fileSize) throws IOException {

-        List<ChunkView> chunkViews = viewFromVisibles(visibleIntervals, position, bufferLength);
+        List<ChunkView> chunkViews = viewFromVisibles(visibleIntervals, position, buf.remaining());

        Map<String, FilerProto.Locations> knownLocations = new HashMap<>();

@@ -59,6 +59,7 @@ public class SeaweedRead {
            if (startOffset < chunkView.logicOffset) {
                long gap = chunkView.logicOffset - startOffset;
                LOG.debug("zero [{},{})", startOffset, startOffset + gap);
+                buf.position(buf.position()+ (int)gap);
                readCount += gap;
                startOffset += gap;
            }
@@ -70,7 +71,7 @@ public class SeaweedRead {
                return 0;
            }

-            int len = readChunkView(startOffset, buffer, bufferOffset + readCount, chunkView, locations);
+            int len = readChunkView(startOffset, buf, chunkView, locations);

            LOG.debug("read [{},{}) {} size {}", startOffset, startOffset + len, chunkView.fileId, chunkView.size);

@@ -79,11 +80,12 @@ public class SeaweedRead {

        }

-        long limit = Math.min(bufferOffset + bufferLength, fileSize);
+        long limit = Math.min(buf.limit(), fileSize);

        if (startOffset < limit) {
            long gap = limit - startOffset;
            LOG.debug("zero2 [{},{})", startOffset, startOffset + gap);
+            buf.position(buf.position()+ (int)gap);
            readCount += gap;
            startOffset += gap;
        }
@@ -91,7 +93,7 @@ public class SeaweedRead {
        return readCount;
    }

-    private static int readChunkView(long startOffset, byte[] buffer, long bufOffset, ChunkView chunkView, FilerProto.Locations locations) throws IOException {
+    private static int readChunkView(long startOffset, ByteBuffer buf, ChunkView chunkView, FilerProto.Locations locations) throws IOException {

        byte[] chunkData = chunkCache.getChunk(chunkView.fileId);

@@ -101,9 +103,9 @@ public class SeaweedRead {
        }

        int len = (int) chunkView.size;
-        LOG.debug("readChunkView fid:{} chunkData.length:{} chunkView.offset:{} chunkView[{};{}) buf[{},{})/{} startOffset:{}",
-                chunkView.fileId, chunkData.length, chunkView.offset, chunkView.logicOffset, chunkView.logicOffset + chunkView.size, bufOffset, bufOffset + len, buffer.length, startOffset);
-        System.arraycopy(chunkData, (int) (startOffset - chunkView.logicOffset + chunkView.offset), buffer, (int) bufOffset, len);
+        LOG.debug("readChunkView fid:{} chunkData.length:{} chunkView.offset:{} chunkView[{};{}) startOffset:{}",
+                chunkView.fileId, chunkData.length, chunkView.offset, chunkView.logicOffset, chunkView.logicOffset + chunkView.size, startOffset);
+        buf.put(chunkData, (int) (startOffset - chunkView.logicOffset + chunkView.offset), len);

        return len;
    }
@@ -100,6 +100,7 @@ message Entry {
    map<string, bytes> extended = 5;
    bytes hard_link_id = 7;
    int32 hard_link_counter = 8; // only exists in hard link meta data
+    bytes content = 9; // if not empty, the file content
 }

 message FullEntry {
@@ -11,13 +11,13 @@
        <dependency>
            <groupId>com.github.chrislusf</groupId>
            <artifactId>seaweedfs-client</artifactId>
-            <version>1.5.6</version>
+            <version>1.5.7</version>
            <scope>compile</scope>
        </dependency>
        <dependency>
            <groupId>com.github.chrislusf</groupId>
            <artifactId>seaweedfs-hadoop2-client</artifactId>
-            <version>1.5.6</version>
+            <version>1.5.7</version>
            <scope>compile</scope>
        </dependency>
        <dependency>
@@ -1,6 +1,6 @@
 package com.seaweedfs.examples;

-import seaweed.hdfs.SeaweedInputStream;
+import seaweed.hdfs.SeaweedHadoopInputStream;
 import seaweedfs.client.FilerClient;
 import seaweedfs.client.FilerGrpcClient;

@@ -24,7 +24,7 @@ public class UnzipFile {

        long localProcessTime = startTime2 - startTime;

-        SeaweedInputStream seaweedInputStream = new SeaweedInputStream(
+        SeaweedHadoopInputStream seaweedInputStream = new SeaweedHadoopInputStream(
                filerGrpcClient,
                new org.apache.hadoop.fs.FileSystem.Statistics(""),
                "/",
@@ -301,7 +301,7 @@
    </snapshotRepository>
  </distributionManagement>
  <properties>
-    <seaweedfs.client.version>1.5.6</seaweedfs.client.version>
+    <seaweedfs.client.version>1.5.7</seaweedfs.client.version>
    <hadoop.version>2.9.2</hadoop.version>
  </properties>
 </project>
@@ -5,7 +5,7 @@
    <modelVersion>4.0.0</modelVersion>

    <properties>
-        <seaweedfs.client.version>1.5.6</seaweedfs.client.version>
+        <seaweedfs.client.version>1.5.7</seaweedfs.client.version>
        <hadoop.version>2.9.2</hadoop.version>
    </properties>

@@ -0,0 +1,25 @@
+package seaweed.hdfs;
+
+import org.apache.hadoop.fs.*;
+
+import java.io.IOException;
+import java.nio.ByteBuffer;
+
+public class BufferedByteBufferReadableInputStream extends BufferedFSInputStream implements ByteBufferReadable {
+
+    public BufferedByteBufferReadableInputStream(FSInputStream in, int size) {
+        super(in, size);
+        if (!(in instanceof Seekable) || !(in instanceof PositionedReadable)) {
+            throw new IllegalArgumentException("In is not an instance of Seekable or PositionedReadable");
+        }
+    }
+
+    @Override
+    public int read(ByteBuffer buf) throws IOException {
+        if (this.in instanceof ByteBufferReadable) {
+            return ((ByteBufferReadable)this.in).read(buf);
+        } else {
+            throw new UnsupportedOperationException("Byte-buffer read unsupported by input stream");
+        }
+    }
+}
@@ -82,7 +82,7 @@ public class SeaweedFileSystem extends FileSystem {
        try {
            int seaweedBufferSize = this.getConf().getInt(FS_SEAWEED_BUFFER_SIZE, FS_SEAWEED_DEFAULT_BUFFER_SIZE);
            FSInputStream inputStream = seaweedFileSystemStore.openFileForRead(path, statistics);
-            return new FSDataInputStream(new BufferedFSInputStream(inputStream, 4 * seaweedBufferSize));
+            return new FSDataInputStream(new BufferedByteBufferReadableInputStream(inputStream, 4 * seaweedBufferSize));
        } catch (Exception ex) {
            LOG.warn("open path: {} bufferSize:{}", path, bufferSize, ex);
            return null;
@@ -193,6 +193,7 @@ public class SeaweedFileSystemStore {
            if (existingEntry != null) {
                entry = FilerProto.Entry.newBuilder();
                entry.mergeFrom(existingEntry);
+                entry.clearContent();
                entry.getAttributesBuilder().setMtime(now);
                LOG.debug("createFile merged entry path:{} entry:{} from:{}", path, entry, existingEntry);
                writePosition = SeaweedRead.fileSize(existingEntry);
@@ -215,7 +216,7 @@ public class SeaweedFileSystemStore {
            SeaweedWrite.writeMeta(filerGrpcClient, getParentDirectory(path), entry);
        }

-        return new SeaweedOutputStream(filerGrpcClient, path, entry, writePosition, bufferSize, replication);
+        return new SeaweedHadoopOutputStream(filerGrpcClient, path.toString(), entry, writePosition, bufferSize, replication);

    }

@@ -229,7 +230,7 @@ public class SeaweedFileSystemStore {
            throw new FileNotFoundException("read non-exist file " + path);
        }

-        return new SeaweedInputStream(filerGrpcClient,
+        return new SeaweedHadoopInputStream(filerGrpcClient,
            statistics,
            path.toUri().getPath(),
            entry);
@@ -0,0 +1,150 @@
+package seaweed.hdfs;
+
+// based on org.apache.hadoop.fs.azurebfs.services.AbfsInputStream
+
+import org.apache.hadoop.fs.ByteBufferReadable;
+import org.apache.hadoop.fs.FSInputStream;
+import org.apache.hadoop.fs.FileSystem.Statistics;
+import seaweedfs.client.FilerGrpcClient;
+import seaweedfs.client.FilerProto;
+import seaweedfs.client.SeaweedInputStream;
+
+import java.io.EOFException;
+import java.io.IOException;
+import java.nio.ByteBuffer;
+
+public class SeaweedHadoopInputStream extends FSInputStream implements ByteBufferReadable {
+
+    private final SeaweedInputStream seaweedInputStream;
+    private final Statistics statistics;
+
+    public SeaweedHadoopInputStream(
+            final FilerGrpcClient filerGrpcClient,
+            final Statistics statistics,
+            final String path,
+            final FilerProto.Entry entry) throws IOException {
+        this.seaweedInputStream = new SeaweedInputStream(filerGrpcClient, path, entry);
+        this.statistics = statistics;
+    }
+
+    @Override
+    public int read() throws IOException {
+        return seaweedInputStream.read();
+    }
+
+    @Override
+    public int read(final byte[] b, final int off, final int len) throws IOException {
+        return seaweedInputStream.read(b, off, len);
+    }
+
+    // implement ByteBufferReadable
+    @Override
+    public synchronized int read(ByteBuffer buf) throws IOException {
+        int bytesRead = seaweedInputStream.read(buf);
+
+        if (bytesRead > 0) {
+            if (statistics != null) {
+                statistics.incrementBytesRead(bytesRead);
+            }
+        }
+
+        return bytesRead;
+    }
+
+    /**
+     * Seek to given position in stream.
+     *
+     * @param n position to seek to
+     * @throws IOException  if there is an error
+     * @throws EOFException if attempting to seek past end of file
+     */
+    @Override
+    public synchronized void seek(long n) throws IOException {
+        seaweedInputStream.seek(n);
+    }
+
+    @Override
+    public synchronized long skip(long n) throws IOException {
+        return seaweedInputStream.skip(n);
+    }
+
+    /**
+     * Return the size of the remaining available bytes
+     * if the size is less than or equal to {@link Integer#MAX_VALUE},
+     * otherwise, return {@link Integer#MAX_VALUE}.
+     * <p>
+     * This is to match the behavior of DFSInputStream.available(),
+     * which some clients may rely on (HBase write-ahead log reading in
+     * particular).
+     */
+    @Override
+    public synchronized int available() throws IOException {
+        return seaweedInputStream.available();
+    }
+
+    /**
+     * Returns the length of the file that this stream refers to. Note that the length returned is the length
+     * as of the time the Stream was opened. Specifically, if there have been subsequent appends to the file,
+     * they wont be reflected in the returned length.
+     *
+     * @return length of the file.
+     * @throws IOException if the stream is closed
+     */
+    public long length() throws IOException {
+        return seaweedInputStream.length();
+    }
+
+    /**
+     * Return the current offset from the start of the file
+     *
+     * @throws IOException throws {@link IOException} if there is an error
+     */
+    @Override
+    public synchronized long getPos() throws IOException {
+        return seaweedInputStream.getPos();
+    }
+
+    /**
+     * Seeks a different copy of the data.  Returns true if
+     * found a new source, false otherwise.
+     *
+     * @throws IOException throws {@link IOException} if there is an error
+     */
+    @Override
+    public boolean seekToNewSource(long l) throws IOException {
+        return false;
+    }
+
+    @Override
+    public synchronized void close() throws IOException {
+        seaweedInputStream.close();
+    }
+
+    /**
+     * Not supported by this stream. Throws {@link UnsupportedOperationException}
+     *
+     * @param readlimit ignored
+     */
+    @Override
+    public synchronized void mark(int readlimit) {
+        throw new UnsupportedOperationException("mark()/reset() not supported on this stream");
+    }
+
+    /**
+     * Not supported by this stream. Throws {@link UnsupportedOperationException}
+     */
+    @Override
+    public synchronized void reset() throws IOException {
+        throw new UnsupportedOperationException("mark()/reset() not supported on this stream");
+    }
+
+    /**
+     * gets whether mark and reset are supported by {@code ADLFileInputStream}. Always returns false.
+     *
+     * @return always {@code false}
+     */
+    @Override
+    public boolean markSupported() {
+        return false;
+    }
+}
@@ -0,0 +1,16 @@
+package seaweed.hdfs;
+
+// adapted from org.apache.hadoop.fs.azurebfs.services.AbfsOutputStream
+
+import seaweedfs.client.FilerGrpcClient;
+import seaweedfs.client.FilerProto;
+import seaweedfs.client.SeaweedOutputStream;
+
+public class SeaweedHadoopOutputStream extends SeaweedOutputStream {
+
+    public SeaweedHadoopOutputStream(FilerGrpcClient filerGrpcClient, final String path, FilerProto.Entry.Builder entry,
+                                     final long position, final int bufferSize, final String replication) {
+        super(filerGrpcClient, path.toString(), entry, position, bufferSize, replication);
+    }
+
+}
@@ -309,7 +309,7 @@
    </snapshotRepository>
  </distributionManagement>
  <properties>
-    <seaweedfs.client.version>1.5.6</seaweedfs.client.version>
+    <seaweedfs.client.version>1.5.7</seaweedfs.client.version>
    <hadoop.version>3.1.1</hadoop.version>
  </properties>
 </project>
@@ -5,7 +5,7 @@
    <modelVersion>4.0.0</modelVersion>

    <properties>
-        <seaweedfs.client.version>1.5.6</seaweedfs.client.version>
+        <seaweedfs.client.version>1.5.7</seaweedfs.client.version>
        <hadoop.version>3.1.1</hadoop.version>
    </properties>

@@ -0,0 +1,25 @@
+package seaweed.hdfs;
+
+import org.apache.hadoop.fs.*;
+
+import java.io.IOException;
+import java.nio.ByteBuffer;
+
+public class BufferedByteBufferReadableInputStream extends BufferedFSInputStream implements ByteBufferReadable {
+
+    public BufferedByteBufferReadableInputStream(FSInputStream in, int size) {
+        super(in, size);
+        if (!(in instanceof Seekable) || !(in instanceof PositionedReadable)) {
+            throw new IllegalArgumentException("In is not an instance of Seekable or PositionedReadable");
+        }
+    }
+
+    @Override
+    public int read(ByteBuffer buf) throws IOException {
+        if (this.in instanceof ByteBufferReadable) {
+            return ((ByteBufferReadable)this.in).read(buf);
+        } else {
+            throw new UnsupportedOperationException("Byte-buffer read unsupported by input stream");
+        }
+    }
+}
@@ -82,7 +82,7 @@ public class SeaweedFileSystem extends FileSystem {
        try {
            int seaweedBufferSize = this.getConf().getInt(FS_SEAWEED_BUFFER_SIZE, FS_SEAWEED_DEFAULT_BUFFER_SIZE);
            FSInputStream inputStream = seaweedFileSystemStore.openFileForRead(path, statistics);
-            return new FSDataInputStream(new BufferedFSInputStream(inputStream, 4 * seaweedBufferSize));
+            return new FSDataInputStream(new BufferedByteBufferReadableInputStream(inputStream, 4 * seaweedBufferSize));
        } catch (Exception ex) {
            LOG.warn("open path: {} bufferSize:{}", path, bufferSize, ex);
            return null;
@@ -193,6 +193,7 @@ public class SeaweedFileSystemStore {
            if (existingEntry != null) {
                entry = FilerProto.Entry.newBuilder();
                entry.mergeFrom(existingEntry);
+                entry.clearContent();
                entry.getAttributesBuilder().setMtime(now);
                LOG.debug("createFile merged entry path:{} entry:{} from:{}", path, entry, existingEntry);
                writePosition = SeaweedRead.fileSize(existingEntry);
@@ -215,7 +216,7 @@ public class SeaweedFileSystemStore {
            SeaweedWrite.writeMeta(filerGrpcClient, getParentDirectory(path), entry);
        }

-        return new SeaweedOutputStream(filerGrpcClient, path, entry, writePosition, bufferSize, replication);
+        return new SeaweedHadoopOutputStream(filerGrpcClient, path.toString(), entry, writePosition, bufferSize, replication);

    }

@@ -229,7 +230,7 @@ public class SeaweedFileSystemStore {
            throw new FileNotFoundException("read non-exist file " + path);
        }

-        return new SeaweedInputStream(filerGrpcClient,
+        return new SeaweedHadoopInputStream(filerGrpcClient,
            statistics,
            path.toUri().getPath(),
            entry);
@@ -287,4 +288,4 @@ public class SeaweedFileSystemStore {

    }

-}
+}
@@ -0,0 +1,150 @@
+package seaweed.hdfs;
+
+// based on org.apache.hadoop.fs.azurebfs.services.AbfsInputStream
+
+import org.apache.hadoop.fs.ByteBufferReadable;
+import org.apache.hadoop.fs.FSInputStream;
+import org.apache.hadoop.fs.FileSystem.Statistics;
+import seaweedfs.client.FilerGrpcClient;
+import seaweedfs.client.FilerProto;
+import seaweedfs.client.SeaweedInputStream;
+
+import java.io.EOFException;
+import java.io.IOException;
+import java.nio.ByteBuffer;
+
+public class SeaweedHadoopInputStream extends FSInputStream implements ByteBufferReadable {
+
+    private final SeaweedInputStream seaweedInputStream;
+    private final Statistics statistics;
+
+    public SeaweedHadoopInputStream(
+            final FilerGrpcClient filerGrpcClient,
+            final Statistics statistics,
+            final String path,
+            final FilerProto.Entry entry) throws IOException {
+        this.seaweedInputStream = new SeaweedInputStream(filerGrpcClient, path, entry);
+        this.statistics = statistics;
+    }
+
+    @Override
+    public int read() throws IOException {
+        return seaweedInputStream.read();
+    }
+
+    @Override
+    public int read(final byte[] b, final int off, final int len) throws IOException {
+        return seaweedInputStream.read(b, off, len);
+    }
+
+    // implement ByteBufferReadable
+    @Override
+    public synchronized int read(ByteBuffer buf) throws IOException {
+        int bytesRead = seaweedInputStream.read(buf);
+
+        if (bytesRead > 0) {
+            if (statistics != null) {
+                statistics.incrementBytesRead(bytesRead);
+            }
+        }
+
+        return bytesRead;
+    }
+
+    /**
+     * Seek to given position in stream.
+     *
+     * @param n position to seek to
+     * @throws IOException  if there is an error
+     * @throws EOFException if attempting to seek past end of file
+     */
+    @Override
+    public synchronized void seek(long n) throws IOException {
+        seaweedInputStream.seek(n);
+    }
+
+    @Override
+    public synchronized long skip(long n) throws IOException {
+        return seaweedInputStream.skip(n);
+    }
+
+    /**
+     * Return the size of the remaining available bytes
+     * if the size is less than or equal to {@link Integer#MAX_VALUE},
+     * otherwise, return {@link Integer#MAX_VALUE}.
+     * <p>
+     * This is to match the behavior of DFSInputStream.available(),
+     * which some clients may rely on (HBase write-ahead log reading in
+     * particular).
+     */
+    @Override
+    public synchronized int available() throws IOException {
+        return seaweedInputStream.available();
+    }
+
+    /**
+     * Returns the length of the file that this stream refers to. Note that the length returned is the length
+     * as of the time the Stream was opened. Specifically, if there have been subsequent appends to the file,
+     * they wont be reflected in the returned length.
+     *
+     * @return length of the file.
+     * @throws IOException if the stream is closed
+     */
+    public long length() throws IOException {
+        return seaweedInputStream.length();
+    }
+
+    /**
+     * Return the current offset from the start of the file
+     *
+     * @throws IOException throws {@link IOException} if there is an error
+     */
+    @Override
+    public synchronized long getPos() throws IOException {
+        return seaweedInputStream.getPos();
+    }
+
+    /**
+     * Seeks a different copy of the data.  Returns true if
+     * found a new source, false otherwise.
+     *
+     * @throws IOException throws {@link IOException} if there is an error
+     */
+    @Override
+    public boolean seekToNewSource(long l) throws IOException {
+        return false;
+    }
+
+    @Override
+    public synchronized void close() throws IOException {
+        seaweedInputStream.close();
+    }
+
+    /**
+     * Not supported by this stream. Throws {@link UnsupportedOperationException}
+     *
+     * @param readlimit ignored
+     */
+    @Override
+    public synchronized void mark(int readlimit) {
+        throw new UnsupportedOperationException("mark()/reset() not supported on this stream");
+    }
+
+    /**
+     * Not supported by this stream. Throws {@link UnsupportedOperationException}
+     */
+    @Override
+    public synchronized void reset() throws IOException {
+        throw new UnsupportedOperationException("mark()/reset() not supported on this stream");
+    }
+
+    /**
+     * gets whether mark and reset are supported by {@code ADLFileInputStream}. Always returns false.
+     *
+     * @return always {@code false}
+     */
+    @Override
+    public boolean markSupported() {
+        return false;
+    }
+}
@@ -0,0 +1,64 @@
+package seaweed.hdfs;
+
+// adapted from org.apache.hadoop.fs.azurebfs.services.AbfsOutputStream
+
+import org.apache.hadoop.fs.StreamCapabilities;
+import org.apache.hadoop.fs.Syncable;
+import seaweedfs.client.FilerGrpcClient;
+import seaweedfs.client.FilerProto;
+import seaweedfs.client.SeaweedOutputStream;
+
+import java.io.IOException;
+import java.util.Locale;
+
+public class SeaweedHadoopOutputStream extends SeaweedOutputStream implements Syncable, StreamCapabilities {
+
+    public SeaweedHadoopOutputStream(FilerGrpcClient filerGrpcClient, final String path, FilerProto.Entry.Builder entry,
+                                     final long position, final int bufferSize, final String replication) {
+        super(filerGrpcClient, path, entry, position, bufferSize, replication);
+    }
+
+    /**
+     * Similar to posix fsync, flush out the data in client's user buffer
+     * all the way to the disk device (but the disk may have it in its cache).
+     *
+     * @throws IOException if error occurs
+     */
+    @Override
+    public void hsync() throws IOException {
+        if (supportFlush) {
+            flushInternal();
+        }
+    }
+
+    /**
+     * Flush out the data in client's user buffer. After the return of
+     * this call, new readers will see the data.
+     *
+     * @throws IOException if any error occurs
+     */
+    @Override
+    public void hflush() throws IOException {
+        if (supportFlush) {
+            flushInternal();
+        }
+    }
+
+    /**
+     * Query the stream for a specific capability.
+     *
+     * @param capability string to query the stream support for.
+     * @return true for hsync and hflush.
+     */
+    @Override
+    public boolean hasCapability(String capability) {
+        switch (capability.toLowerCase(Locale.ENGLISH)) {
+            case StreamCapabilities.HSYNC:
+            case StreamCapabilities.HFLUSH:
+                return supportFlush;
+            default:
+                return false;
+        }
+    }
+
+}
@@ -1,240 +0,0 @@
-package seaweed.hdfs;
-
-// based on org.apache.hadoop.fs.azurebfs.services.AbfsInputStream
-
-import org.apache.hadoop.fs.FSExceptionMessages;
-import org.apache.hadoop.fs.FSInputStream;
-import org.apache.hadoop.fs.FileSystem.Statistics;
-import org.slf4j.Logger;
-import org.slf4j.LoggerFactory;
-import seaweedfs.client.FilerGrpcClient;
-import seaweedfs.client.FilerProto;
-import seaweedfs.client.SeaweedRead;
-
-import java.io.EOFException;
-import java.io.IOException;
-import java.util.List;
-
-public class SeaweedInputStream extends FSInputStream {
-
-    private static final Logger LOG = LoggerFactory.getLogger(SeaweedInputStream.class);
-
-    private final FilerGrpcClient filerGrpcClient;
-    private final Statistics statistics;
-    private final String path;
-    private final FilerProto.Entry entry;
-    private final List<SeaweedRead.VisibleInterval> visibleIntervalList;
-    private final long contentLength;
-
-    private long position = 0;  // cursor of the file
-
-    private boolean closed = false;
-
-    public SeaweedInputStream(
-            final FilerGrpcClient filerGrpcClient,
-            final Statistics statistics,
-            final String path,
-            final FilerProto.Entry entry) throws IOException {
-        this.filerGrpcClient = filerGrpcClient;
-        this.statistics = statistics;
-        this.path = path;
-        this.entry = entry;
-        this.contentLength = SeaweedRead.fileSize(entry);
-
-        this.visibleIntervalList = SeaweedRead.nonOverlappingVisibleIntervals(filerGrpcClient, entry.getChunksList());
-
-        LOG.debug("new path:{} entry:{} visibleIntervalList:{}", path, entry, visibleIntervalList);
-
-    }
-
-    public String getPath() {
-        return path;
-    }
-
-    @Override
-    public int read() throws IOException {
-        byte[] b = new byte[1];
-        int numberOfBytesRead = read(b, 0, 1);
-        if (numberOfBytesRead < 0) {
-            return -1;
-        } else {
-            return (b[0] & 0xFF);
-        }
-    }
-
-    @Override
-    public synchronized int read(final byte[] b, final int off, final int len) throws IOException {
-
-        if (position < 0) {
-            throw new IllegalArgumentException("attempting to read from negative offset");
-        }
-        if (position >= contentLength) {
-            return -1;  // Hadoop prefers -1 to EOFException
-        }
-        if (b == null) {
-            throw new IllegalArgumentException("null byte array passed in to read() method");
-        }
-        if (off >= b.length) {
-            throw new IllegalArgumentException("offset greater than length of array");
-        }
-        if (len < 0) {
-            throw new IllegalArgumentException("requested read length is less than zero");
-        }
-        if (len > (b.length - off)) {
-            throw new IllegalArgumentException("requested read length is more than will fit after requested offset in buffer");
-        }
-
-        long bytesRead = SeaweedRead.read(this.filerGrpcClient, this.visibleIntervalList, this.position, b, off, len, SeaweedRead.fileSize(entry));
-        if (bytesRead > Integer.MAX_VALUE) {
-            throw new IOException("Unexpected Content-Length");
-        }
-
-        if (bytesRead > 0) {
-            this.position += bytesRead;
-            if (statistics != null) {
-                statistics.incrementBytesRead(bytesRead);
-            }
-        }
-
-        return (int) bytesRead;
-
-    }
-
-    /**
-     * Seek to given position in stream.
-     *
-     * @param n position to seek to
-     * @throws IOException  if there is an error
-     * @throws EOFException if attempting to seek past end of file
-     */
-    @Override
-    public synchronized void seek(long n) throws IOException {
-        if (closed) {
-            throw new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
-        }
-        if (n < 0) {
-            throw new EOFException(FSExceptionMessages.NEGATIVE_SEEK);
-        }
-        if (n > contentLength) {
-            throw new EOFException(FSExceptionMessages.CANNOT_SEEK_PAST_EOF);
-        }
-
-        this.position = n;
-
-    }
-
-    @Override
-    public synchronized long skip(long n) throws IOException {
-        if (closed) {
-            throw new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
-        }
-        if (this.position == contentLength) {
-            if (n > 0) {
-                throw new EOFException(FSExceptionMessages.CANNOT_SEEK_PAST_EOF);
-            }
-        }
-        long newPos = this.position + n;
-        if (newPos < 0) {
-            newPos = 0;
-            n = newPos - this.position;
-        }
-        if (newPos > contentLength) {
-            newPos = contentLength;
-            n = newPos - this.position;
-        }
-        seek(newPos);
-        return n;
-    }
-
-    /**
-     * Return the size of the remaining available bytes
-     * if the size is less than or equal to {@link Integer#MAX_VALUE},
-     * otherwise, return {@link Integer#MAX_VALUE}.
-     * <p>
-     * This is to match the behavior of DFSInputStream.available(),
-     * which some clients may rely on (HBase write-ahead log reading in
-     * particular).
-     */
-    @Override
-    public synchronized int available() throws IOException {
-        if (closed) {
-            throw new IOException(
-                    FSExceptionMessages.STREAM_IS_CLOSED);
-        }
-        final long remaining = this.contentLength - this.getPos();
-        return remaining <= Integer.MAX_VALUE
-                ? (int) remaining : Integer.MAX_VALUE;
-    }
-
-    /**
-     * Returns the length of the file that this stream refers to. Note that the length returned is the length
-     * as of the time the Stream was opened. Specifically, if there have been subsequent appends to the file,
-     * they wont be reflected in the returned length.
-     *
-     * @return length of the file.
-     * @throws IOException if the stream is closed
-     */
-    public long length() throws IOException {
-        if (closed) {
-            throw new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
-        }
-        return contentLength;
-    }
-
-    /**
-     * Return the current offset from the start of the file
-     *
-     * @throws IOException throws {@link IOException} if there is an error
-     */
-    @Override
-    public synchronized long getPos() throws IOException {
-        if (closed) {
-            throw new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
-        }
-        return position;
-    }
-
-    /**
-     * Seeks a different copy of the data.  Returns true if
-     * found a new source, false otherwise.
-     *
-     * @throws IOException throws {@link IOException} if there is an error
-     */
-    @Override
-    public boolean seekToNewSource(long l) throws IOException {
-        return false;
-    }
-
-    @Override
-    public synchronized void close() throws IOException {
-        closed = true;
-    }
-
-    /**
-     * Not supported by this stream. Throws {@link UnsupportedOperationException}
-     *
-     * @param readlimit ignored
-     */
-    @Override
-    public synchronized void mark(int readlimit) {
-        throw new UnsupportedOperationException("mark()/reset() not supported on this stream");
-    }
-
-    /**
-     * Not supported by this stream. Throws {@link UnsupportedOperationException}
-     */
-    @Override
-    public synchronized void reset() throws IOException {
-        throw new UnsupportedOperationException("mark()/reset() not supported on this stream");
-    }
-
-    /**
-     * gets whether mark and reset are supported by {@code ADLFileInputStream}. Always returns false.
-     *
-     * @return always {@code false}
-     */
-    @Override
-    public boolean markSupported() {
-        return false;
-    }
-}
@@ -1,337 +0,0 @@
-package seaweed.hdfs;
-
-// adapted from org.apache.hadoop.fs.azurebfs.services.AbfsOutputStream
-
-import com.google.common.base.Preconditions;
-import org.apache.hadoop.fs.FSExceptionMessages;
-import org.apache.hadoop.fs.Path;
-import org.apache.hadoop.fs.StreamCapabilities;
-import org.apache.hadoop.fs.Syncable;
-import org.slf4j.Logger;
-import org.slf4j.LoggerFactory;
-import seaweedfs.client.ByteBufferPool;
-import seaweedfs.client.FilerGrpcClient;
-import seaweedfs.client.FilerProto;
-import seaweedfs.client.SeaweedWrite;
-
-import java.io.IOException;
-import java.io.InterruptedIOException;
-import java.io.OutputStream;
-import java.nio.ByteBuffer;
-import java.util.Locale;
-import java.util.concurrent.*;
-
-import static seaweed.hdfs.SeaweedFileSystemStore.getParentDirectory;
-
-public class SeaweedOutputStream extends OutputStream implements Syncable, StreamCapabilities {
-
-    private static final Logger LOG = LoggerFactory.getLogger(SeaweedOutputStream.class);
-
-    private final FilerGrpcClient filerGrpcClient;
-    private final Path path;
-    private final int bufferSize;
-    private final int maxConcurrentRequestCount;
-    private final ThreadPoolExecutor threadExecutor;
-    private final ExecutorCompletionService<Void> completionService;
-    private final FilerProto.Entry.Builder entry;
-    private final boolean supportFlush = false; // true;
-    private final ConcurrentLinkedDeque<WriteOperation> writeOperations;
-    private long position;
-    private boolean closed;
-    private volatile IOException lastError;
-    private long lastFlushOffset;
-    private long lastTotalAppendOffset = 0;
-    private ByteBuffer buffer;
-    private long outputIndex;
-    private String replication = "000";
-
-    public SeaweedOutputStream(FilerGrpcClient filerGrpcClient, final Path path, FilerProto.Entry.Builder entry,
-                               final long position, final int bufferSize, final String replication) {
-        this.filerGrpcClient = filerGrpcClient;
-        this.replication = replication;
-        this.path = path;
-        this.position = position;
-        this.closed = false;
-        this.lastError = null;
-        this.lastFlushOffset = 0;
-        this.bufferSize = bufferSize;
-        this.buffer = ByteBufferPool.request(bufferSize);
-        this.outputIndex = 0;
-        this.writeOperations = new ConcurrentLinkedDeque<>();
-
-        this.maxConcurrentRequestCount = Runtime.getRuntime().availableProcessors();
-
-        this.threadExecutor
-                = new ThreadPoolExecutor(maxConcurrentRequestCount,
-                maxConcurrentRequestCount,
-                120L,
-                TimeUnit.SECONDS,
-                new LinkedBlockingQueue<Runnable>());
-        this.completionService = new ExecutorCompletionService<>(this.threadExecutor);
-
-        this.entry = entry;
-
-    }
-
-    private synchronized void flushWrittenBytesToServiceInternal(final long offset) throws IOException {
-        try {
-            SeaweedWrite.writeMeta(filerGrpcClient, getParentDirectory(path), entry);
-        } catch (Exception ex) {
-            throw new IOException(ex);
-        }
-        this.lastFlushOffset = offset;
-    }
-
-    @Override
-    public void write(final int byteVal) throws IOException {
-        write(new byte[]{(byte) (byteVal & 0xFF)});
-    }
-
-    @Override
-    public synchronized void write(final byte[] data, final int off, final int length)
-            throws IOException {
-        maybeThrowLastError();
-
-        Preconditions.checkArgument(data != null, "null data");
-
-        if (off < 0 || length < 0 || length > data.length - off) {
-            throw new IndexOutOfBoundsException();
-        }
-
-        // System.out.println(path + " write [" + (outputIndex + off) + "," + ((outputIndex + off) + length) + ")");
-
-        int currentOffset = off;
-        int writableBytes = bufferSize - buffer.position();
-        int numberOfBytesToWrite = length;
-
-        while (numberOfBytesToWrite > 0) {
-
-            if (numberOfBytesToWrite < writableBytes) {
-                buffer.put(data, currentOffset, numberOfBytesToWrite);
-                outputIndex += numberOfBytesToWrite;
-                break;
-            }
-
-            // System.out.println(path + "     [" + (outputIndex + currentOffset) + "," + ((outputIndex + currentOffset) + writableBytes) + ") " + buffer.capacity());
-            buffer.put(data, currentOffset, writableBytes);
-            outputIndex += writableBytes;
-            currentOffset += writableBytes;
-            writeCurrentBufferToService();
-            numberOfBytesToWrite = numberOfBytesToWrite - writableBytes;
-            writableBytes = bufferSize - buffer.position();
-        }
-
-    }
-
-    /**
-     * Flushes this output stream and forces any buffered output bytes to be
-     * written out. If any data remains in the payload it is committed to the
-     * service. Data is queued for writing and forced out to the service
-     * before the call returns.
-     */
-    @Override
-    public void flush() throws IOException {
-        if (supportFlush) {
-            flushInternalAsync();
-        }
-    }
-
-    /**
-     * Similar to posix fsync, flush out the data in client's user buffer
-     * all the way to the disk device (but the disk may have it in its cache).
-     *
-     * @throws IOException if error occurs
-     */
-    @Override
-    public void hsync() throws IOException {
-        if (supportFlush) {
-            flushInternal();
-        }
-    }
-
-    /**
-     * Flush out the data in client's user buffer. After the return of
-     * this call, new readers will see the data.
-     *
-     * @throws IOException if any error occurs
-     */
-    @Override
-    public void hflush() throws IOException {
-        if (supportFlush) {
-            flushInternal();
-        }
-    }
-
-    /**
-     * Query the stream for a specific capability.
-     *
-     * @param capability string to query the stream support for.
-     * @return true for hsync and hflush.
-     */
-    @Override
-    public boolean hasCapability(String capability) {
-        switch (capability.toLowerCase(Locale.ENGLISH)) {
-            case StreamCapabilities.HSYNC:
-            case StreamCapabilities.HFLUSH:
-                return supportFlush;
-            default:
-                return false;
-        }
-    }
-
-    /**
-     * Force all data in the output stream to be written to Azure storage.
-     * Wait to return until this is complete. Close the access to the stream and
-     * shutdown the upload thread pool.
-     * If the blob was created, its lease will be released.
-     * Any error encountered caught in threads and stored will be rethrown here
-     * after cleanup.
-     */
-    @Override
-    public synchronized void close() throws IOException {
-        if (closed) {
-            return;
-        }
-
-        LOG.debug("close path: {}", path);
-        try {
-            flushInternal();
-            threadExecutor.shutdown();
-        } finally {
-            lastError = new IOException(FSExceptionMessages.STREAM_IS_CLOSED);
-            ByteBufferPool.release(buffer);
-            buffer = null;
-            outputIndex = 0;
-            closed = true;
-            writeOperations.clear();
-            if (!threadExecutor.isShutdown()) {
-                threadExecutor.shutdownNow();
-            }
-        }
-    }
-
-    private synchronized void writeCurrentBufferToService() throws IOException {
-        if (buffer.position() == 0) {
-            return;
-        }
-
-        position += submitWriteBufferToService(buffer, position);
-
-        buffer = ByteBufferPool.request(bufferSize);
-
-    }
-
-    private synchronized int submitWriteBufferToService(final ByteBuffer bufferToWrite, final long writePosition) throws IOException {
-
-        bufferToWrite.flip();
-        int bytesLength = bufferToWrite.limit() - bufferToWrite.position();
-
-        if (threadExecutor.getQueue().size() >= maxConcurrentRequestCount) {
-            waitForTaskToComplete();
-        }
-        final Future<Void> job = completionService.submit(() -> {
-            // System.out.println(path + " is going to save [" + (writePosition) + "," + ((writePosition) + bytesLength) + ")");
-            SeaweedWrite.writeData(entry, replication, filerGrpcClient, writePosition, bufferToWrite.array(), bufferToWrite.position(), bufferToWrite.limit(), path.toUri().getPath());
-            // System.out.println(path + " saved [" + (writePosition) + "," + ((writePosition) + bytesLength) + ")");
-            ByteBufferPool.release(bufferToWrite);
-            return null;
-        });
-
-        writeOperations.add(new WriteOperation(job, writePosition, bytesLength));
-
-        // Try to shrink the queue
-        shrinkWriteOperationQueue();
-
-        return bytesLength;
-
-    }
-
-    private void waitForTaskToComplete() throws IOException {
-        boolean completed;
-        for (completed = false; completionService.poll() != null; completed = true) {
-            // keep polling until there is no data
-        }
-
-        if (!completed) {
-            try {
-                completionService.take();
-            } catch (InterruptedException e) {
-                lastError = (IOException) new InterruptedIOException(e.toString()).initCause(e);
-                throw lastError;
-            }
-        }
-    }
-
-    private void maybeThrowLastError() throws IOException {
-        if (lastError != null) {
-            throw lastError;
-        }
-    }
-
-    /**
-     * Try to remove the completed write operations from the beginning of write
-     * operation FIFO queue.
-     */
-    private synchronized void shrinkWriteOperationQueue() throws IOException {
-        try {
-            while (writeOperations.peek() != null && writeOperations.peek().task.isDone()) {
-                writeOperations.peek().task.get();
-                lastTotalAppendOffset += writeOperations.peek().length;
-                writeOperations.remove();
-            }
-        } catch (Exception e) {
-            lastError = new IOException(e);
-            throw lastError;
-        }
-    }
-
-    private synchronized void flushInternal() throws IOException {
-        maybeThrowLastError();
-        writeCurrentBufferToService();
-        flushWrittenBytesToService();
-    }
-
-    private synchronized void flushInternalAsync() throws IOException {
-        maybeThrowLastError();
-        writeCurrentBufferToService();
-        flushWrittenBytesToServiceAsync();
-    }
-
-    private synchronized void flushWrittenBytesToService() throws IOException {
-        for (WriteOperation writeOperation : writeOperations) {
-            try {
-                writeOperation.task.get();
-            } catch (Exception ex) {
-                lastError = new IOException(ex);
-                throw lastError;
-            }
-        }
-        LOG.debug("flushWrittenBytesToService: {} position:{}", path, position);
-        flushWrittenBytesToServiceInternal(position);
-    }
-
-    private synchronized void flushWrittenBytesToServiceAsync() throws IOException {
-        shrinkWriteOperationQueue();
-
-        if (this.lastTotalAppendOffset > this.lastFlushOffset) {
-            this.flushWrittenBytesToServiceInternal(this.lastTotalAppendOffset);
-        }
-    }
-
-    private static class WriteOperation {
-        private final Future<Void> task;
-        private final long startOffset;
-        private final long length;
-
-        WriteOperation(final Future<Void> task, final long startOffset, final long length) {
-            Preconditions.checkNotNull(task, "task");
-            Preconditions.checkArgument(startOffset >= 0, "startOffset");
-            Preconditions.checkArgument(length >= 0, "length");
-
-            this.task = task;
-            this.startOffset = startOffset;
-            this.length = length;
-        }
-    }
-
-}
@@ -386,7 +386,7 @@
          "steppedLine": false,
          "targets": [
            {
-              "expr": "rate(SeaweedFS_filer_request_total[1m]) * 5",
+              "expr": "rate(SeaweedFS_filer_request_total[1m])",
              "format": "time_series",
              "intervalFactor": 2,
              "legendFormat": "{{type}}",
@@ -761,7 +761,7 @@
          "steppedLine": false,
          "targets": [
            {
-              "expr": "rate(SeaweedFS_s3_request_total[1m]) * 5",
+              "expr": "rate(SeaweedFS_s3_request_total[1m])",
              "format": "time_series",
              "intervalFactor": 2,
              "legendFormat": "{{type}}",
@@ -1853,4 +1853,4 @@
  "timezone": "browser",
  "title": "SeaweedFS",
  "version": 2
-}
+}
@@ -97,7 +97,7 @@
      "steppedLine": false,
      "targets": [
        {
-          "expr": "sum(rate(SeaweedFS_s3_request_total{namespace=\"$namespace\",service=~\"$service-api\",type=~\"$method\"}[1m]) * 5) by (code)",
+          "expr": "sum(rate(SeaweedFS_s3_request_total{namespace=\"$namespace\",service=~\"$service-api\",type=~\"$method\"}[1m])) by (code)",
          "format": "time_series",
          "interval": "",
          "intervalFactor": 2,
@@ -210,7 +210,7 @@
      "steppedLine": false,
      "targets": [
        {
-          "expr": "sum(rate(SeaweedFS_s3_request_total{namespace=\"$namespace\",service=~\"$service-api\"}[1m]) * 5) by (type)",
+          "expr": "sum(rate(SeaweedFS_s3_request_total{namespace=\"$namespace\",service=~\"$service-api\"}[1m])) by (type)",
          "format": "time_series",
          "interval": "",
          "intervalFactor": 2,
@@ -1014,7 +1014,7 @@
      "steppedLine": false,
      "targets": [
        {
-          "expr": "rate(SeaweedFS_filer_request_total{namespace=\"$namespace\",service=~\"$service-api\"}[1m]) * 5",
+          "expr": "rate(SeaweedFS_filer_request_total{namespace=\"$namespace\",service=~\"$service-api\"}[1m])",
          "format": "time_series",
          "interval": "",
          "intervalFactor": 2,
@@ -2359,4 +2359,4 @@
  "timezone": "browser",
  "title": "SeaweedFS",
  "version": 2
-}
+}
@@ -20,7 +20,7 @@ debug_mount:

 debug_server:
 	go build -gcflags="all=-N -l"
-	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- server -dir=/Volumes/mobile_disk/99 -filer -volume.port=8343 -s3 -volume.max=0
+	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- server -dir=/Volumes/mobile_disk/99 -filer -volume.port=8343 -s3 -volume.max=0 -master.volumeSizeLimitMB=1024 -volume.preStopSeconds=1

 debug_volume:
 	go build -gcflags="all=-N -l"
@@ -29,3 +29,7 @@ debug_volume:
 debug_webdav:
 	go build -gcflags="all=-N -l"
 	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- -v=4 webdav
+
+debug_s3:
+	go build -gcflags="all=-N -l"
+	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- -v=4 s3
@@ -112,7 +112,7 @@ func runBackup(cmd *Command, args []string) bool {
 			return true
 		}
 	}
-	v, err := storage.NewVolume(util.ResolvePath(*s.dir), *s.collection, vid, storage.NeedleMapInMemory, replication, ttl, 0, 0)
+	v, err := storage.NewVolume(util.ResolvePath(*s.dir), util.ResolvePath(*s.dir), *s.collection, vid, storage.NeedleMapInMemory, replication, ttl, 0, 0)
 	if err != nil {
 		fmt.Printf("Error creating or reading from volume %d: %v\n", vid, err)
 		return true
@@ -137,7 +137,7 @@ func runBackup(cmd *Command, args []string) bool {
 		// remove the old data
 		v.Destroy()
 		// recreate an empty volume
-		v, err = storage.NewVolume(util.ResolvePath(*s.dir), *s.collection, vid, storage.NeedleMapInMemory, replication, ttl, 0, 0)
+		v, err = storage.NewVolume(util.ResolvePath(*s.dir), util.ResolvePath(*s.dir), *s.collection, vid, storage.NeedleMapInMemory, replication, ttl, 0, 0)
 		if err != nil {
 			fmt.Printf("Error creating or reading from volume %d: %v\n", vid, err)
 			return true
@@ -15,6 +15,8 @@ var Commands = []*Command{
 	cmdDownload,
 	cmdExport,
 	cmdFiler,
+	cmdFilerCat,
+	cmdFilerMetaTail,
 	cmdFilerReplicate,
 	cmdFilerSynchronize,
 	cmdFix,
@@ -25,7 +27,6 @@ var Commands = []*Command{
 	cmdScaffold,
 	cmdServer,
 	cmdShell,
-	cmdWatch,
 	cmdUpload,
 	cmdVersion,
 	cmdVolume,
@@ -41,8 +41,7 @@ func runCompact(cmd *Command, args []string) bool {
 	preallocate := *compactVolumePreallocate * (1 << 20)

 	vid := needle.VolumeId(*compactVolumeId)
-	v, err := storage.NewVolume(util.ResolvePath(*compactVolumePath), *compactVolumeCollection, vid,
-		storage.NeedleMapInMemory, nil, nil, preallocate, 0)
+	v, err := storage.NewVolume(util.ResolvePath(*compactVolumePath), util.ResolvePath(*compactVolumePath), *compactVolumeCollection, vid, storage.NeedleMapInMemory, nil, nil, preallocate, 0)
 	if err != nil {
 		glog.Fatalf("Load Volume [ERROR] %s\n", err)
 	}
@@ -42,8 +42,7 @@ type FilerOptions struct {
 	cipher                  *bool
 	peers                   *string
 	metricsHttpPort         *int
-
-	// default leveldb directory, used in "weed server" mode
+	saveToFilerLimit        *int
 	defaultLevelDbDirectory *string
 }

@@ -65,6 +64,8 @@ func init() {
 	f.cipher = cmdFiler.Flag.Bool("encryptVolumeData", false, "encrypt data on volume servers")
 	f.peers = cmdFiler.Flag.String("peers", "", "all filers sharing the same filer store in comma separated ip:port list")
 	f.metricsHttpPort = cmdFiler.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
+	f.saveToFilerLimit = cmdFiler.Flag.Int("saveToFilerLimit", 0, "files smaller than this limit will be saved in filer store")
+	f.defaultLevelDbDirectory = cmdFiler.Flag.String("defaultStoreDir", ".", "if filer.toml is empty, use an embedded filer store in the directory")

 	// start s3 on filer
 	filerStartS3 = cmdFiler.Flag.Bool("s3", false, "whether to start S3 gateway")
@@ -73,6 +74,7 @@ func init() {
 	filerS3Options.tlsPrivateKey = cmdFiler.Flag.String("s3.key.file", "", "path to the TLS private key file")
 	filerS3Options.tlsCertificate = cmdFiler.Flag.String("s3.cert.file", "", "path to the TLS certificate file")
 	filerS3Options.config = cmdFiler.Flag.String("s3.config", "", "path to the config file")
+	filerS3Options.allowEmptyFolder = cmdFiler.Flag.Bool("s3.allowEmptyFolder", false, "allow empty folders")
 }

 var cmdFiler = &Command{
@@ -89,7 +91,8 @@ var cmdFiler = &Command{
 	//return a json format subdirectory and files listing
 	GET /path/to/

-	The configuration file "filer.toml" is read from ".", "$HOME/.seaweedfs/", or "/etc/seaweedfs/", in that order.
+	The configuration file "filer.toml" is read from ".", "$HOME/.seaweedfs/", "/usr/local/etc/seaweedfs/", or "/etc/seaweedfs/", in that order.
+	If the "filer.toml" is not found, an embedded filer store will be craeted under "-defaultStoreDir".

 	The example filer.toml configuration file can be generated by "weed scaffold -config=filer"

@@ -125,10 +128,7 @@ func (fo *FilerOptions) startFiler() {
 		publicVolumeMux = http.NewServeMux()
 	}

-	defaultLevelDbDirectory := "./filerldb2"
-	if fo.defaultLevelDbDirectory != nil {
-		defaultLevelDbDirectory = util.ResolvePath(*fo.defaultLevelDbDirectory + "/filerldb2")
-	}
+	defaultLevelDbDirectory := util.ResolvePath(*fo.defaultLevelDbDirectory + "/filerldb2")

 	var peers []string
 	if *fo.peers != "" {
@@ -149,6 +149,7 @@ func (fo *FilerOptions) startFiler() {
 		Host:               *fo.ip,
 		Port:               uint32(*fo.port),
 		Cipher:             *fo.cipher,
+		SaveToFilerLimit:   *fo.saveToFilerLimit,
 		Filers:             peers,
 	})
 	if nfs_err != nil {
@@ -0,0 +1,118 @@
+package command
+
+import (
+	"context"
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/wdclient"
+	"google.golang.org/grpc"
+	"math"
+	"net/url"
+	"os"
+	"strings"
+
+	"github.com/chrislusf/seaweedfs/weed/security"
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+var (
+	filerCat FilerCatOptions
+)
+
+type FilerCatOptions struct {
+	grpcDialOption grpc.DialOption
+	filerAddress   string
+	filerClient    filer_pb.SeaweedFilerClient
+	output         *string
+}
+
+func (fco *FilerCatOptions) GetLookupFileIdFunction() wdclient.LookupFileIdFunctionType {
+	return func(fileId string) (targetUrls []string, err error) {
+		vid := filer.VolumeId(fileId)
+		resp, err := fco.filerClient.LookupVolume(context.Background(), &filer_pb.LookupVolumeRequest{
+			VolumeIds: []string{vid},
+		})
+		if err != nil {
+			return nil, err
+		}
+		locations := resp.LocationsMap[vid]
+		for _, loc := range locations.Locations {
+			targetUrls = append(targetUrls, fmt.Sprintf("http://%s/%s", loc.Url, fileId))
+		}
+		return
+	}
+}
+
+func init() {
+	cmdFilerCat.Run = runFilerCat // break init cycle
+	filerCat.output = cmdFilerCat.Flag.String("o", "", "write to file instead of stdout")
+}
+
+var cmdFilerCat = &Command{
+	UsageLine: "filer.cat [-o <file>] http://localhost:8888/path/to/file",
+	Short:     "copy one file to local",
+	Long: `read one file to stdout or write to a file
+
+`,
+}
+
+func runFilerCat(cmd *Command, args []string) bool {
+
+	util.LoadConfiguration("security", false)
+
+	if len(args) == 0 {
+		return false
+	}
+	filerSource := args[len(args)-1]
+
+	filerUrl, err := url.Parse(filerSource)
+	if err != nil {
+		fmt.Printf("The last argument should be a URL on filer: %v\n", err)
+		return false
+	}
+	urlPath := filerUrl.Path
+	if strings.HasSuffix(urlPath, "/") {
+		fmt.Printf("The last argument should be a file: %v\n", err)
+		return false
+	}
+
+	filerCat.filerAddress = filerUrl.Host
+	filerCat.grpcDialOption = security.LoadClientTLS(util.GetViper(), "grpc.client")
+
+	dir, name := util.FullPath(urlPath).DirAndName()
+
+	writer := os.Stdout
+	if *filerCat.output != "" {
+
+		fmt.Printf("saving %s to %s\n", filerSource, *filerCat.output)
+
+		f, err := os.OpenFile(*filerCat.output, os.O_WRONLY|os.O_CREATE|os.O_TRUNC, 0755)
+		if err != nil {
+			fmt.Printf("open file %s: %v\n", *filerCat.output, err)
+			return false
+		}
+		defer f.Close()
+		writer = f
+	}
+
+	pb.WithFilerClient(filerCat.filerAddress, filerCat.grpcDialOption, func(client filer_pb.SeaweedFilerClient) error {
+
+		request := &filer_pb.LookupDirectoryEntryRequest{
+			Name:      name,
+			Directory: dir,
+		}
+		respLookupEntry, err := filer_pb.LookupEntry(client, request)
+		if err != nil {
+			return err
+		}
+
+		filerCat.filerClient = client
+
+		return filer.StreamContent(&filerCat, writer, respLookupEntry.Entry.Chunks, 0, math.MaxInt64)
+
+	})
+
+	return true
+}
@@ -92,7 +92,7 @@ func runCopy(cmd *Command, args []string) bool {
 	}
 	urlPath := filerUrl.Path
 	if !strings.HasSuffix(urlPath, "/") {
-		fmt.Printf("The last argument should be a folder and end with \"/\": %v\n", err)
+		fmt.Printf("The last argument should be a folder and end with \"/\"\n")
 		return false
 	}

@@ -157,7 +157,7 @@ func runCopy(cmd *Command, args []string) bool {
 		defer close(fileCopyTaskChan)
 		for _, fileOrDir := range fileOrDirs {
 			if err := genFileCopyTask(fileOrDir, urlPath, fileCopyTaskChan); err != nil {
-				fmt.Fprintf(os.Stderr, "gen file list error: %v\n", err)
+				fmt.Fprintf(os.Stderr, "genFileCopyTask : %v\n", err)
 				break
 			}
 		}
@@ -200,7 +200,7 @@ func genFileCopyTask(fileOrDir string, destPath string, fileCopyTaskChan chan Fi

 	fi, err := os.Stat(fileOrDir)
 	if err != nil {
-		fmt.Fprintf(os.Stderr, "Failed to get stat for file %s: %v\n", fileOrDir, err)
+		fmt.Fprintf(os.Stderr, "Error: read file %s: %v\n", fileOrDir, err)
 		return nil
 	}

@@ -0,0 +1,211 @@
+package command
+
+import (
+	"context"
+	"fmt"
+	"github.com/golang/protobuf/jsonpb"
+	jsoniter "github.com/json-iterator/go"
+	"github.com/olivere/elastic/v7"
+	"io"
+	"os"
+	"path/filepath"
+	"strings"
+	"time"
+
+	"github.com/chrislusf/seaweedfs/weed/pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/security"
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+func init() {
+	cmdFilerMetaTail.Run = runFilerMetaTail // break init cycle
+}
+
+var cmdFilerMetaTail = &Command{
+	UsageLine: "filer.meta.tail [-filer=localhost:8888] [-target=/]",
+	Short:     "see recent changes on a filer",
+	Long: `See recent changes on a filer.
+
+	weed filer.meta.tail -timeAgo=30h | grep truncate
+	weed filer.meta.tail -timeAgo=30h | jq .
+	weed filer.meta.tail -timeAgo=30h | jq .eventNotification.newEntry.name
+
+  `,
+}
+
+var (
+	tailFiler   = cmdFilerMetaTail.Flag.String("filer", "localhost:8888", "filer hostname:port")
+	tailTarget  = cmdFilerMetaTail.Flag.String("pathPrefix", "/", "path to a folder or file, or common prefix for the folders or files on filer")
+	tailStart   = cmdFilerMetaTail.Flag.Duration("timeAgo", 0, "start time before now. \"300ms\", \"1.5h\" or \"2h45m\". Valid time units are \"ns\", \"us\" (or \"µs\"), \"ms\", \"s\", \"m\", \"h\"")
+	tailPattern = cmdFilerMetaTail.Flag.String("pattern", "", "full path or just filename pattern, ex: \"/home/?opher\", \"*.pdf\", see https://golang.org/pkg/path/filepath/#Match ")
+	esServers   = cmdFilerMetaTail.Flag.String("es", "", "comma-separated elastic servers http://<host:port>")
+	esIndex     = cmdFilerMetaTail.Flag.String("es.index", "seaweedfs", "ES index name")
+)
+
+func runFilerMetaTail(cmd *Command, args []string) bool {
+
+	grpcDialOption := security.LoadClientTLS(util.GetViper(), "grpc.client")
+
+	var filterFunc func(dir, fname string) bool
+	if *tailPattern != "" {
+		if strings.Contains(*tailPattern, "/") {
+			println("watch path pattern", *tailPattern)
+			filterFunc = func(dir, fname string) bool {
+				matched, err := filepath.Match(*tailPattern, dir+"/"+fname)
+				if err != nil {
+					fmt.Printf("error: %v", err)
+				}
+				return matched
+			}
+		} else {
+			println("watch file pattern", *tailPattern)
+			filterFunc = func(dir, fname string) bool {
+				matched, err := filepath.Match(*tailPattern, fname)
+				if err != nil {
+					fmt.Printf("error: %v", err)
+				}
+				return matched
+			}
+		}
+	}
+
+	shouldPrint := func(resp *filer_pb.SubscribeMetadataResponse) bool {
+		if filterFunc == nil {
+			return true
+		}
+		if resp.EventNotification.OldEntry == nil && resp.EventNotification.NewEntry == nil {
+			return false
+		}
+		if resp.EventNotification.OldEntry != nil && filterFunc(resp.Directory, resp.EventNotification.OldEntry.Name) {
+			return true
+		}
+		if resp.EventNotification.NewEntry != nil && filterFunc(resp.EventNotification.NewParentPath, resp.EventNotification.NewEntry.Name) {
+			return true
+		}
+		return false
+	}
+
+	jsonpbMarshaler := jsonpb.Marshaler{
+		EmitDefaults: false,
+	}
+	eachEntryFunc := func(resp *filer_pb.SubscribeMetadataResponse) error {
+		jsonpbMarshaler.Marshal(os.Stdout, resp)
+		fmt.Fprintln(os.Stdout)
+		return nil
+	}
+	if *esServers != "" {
+		var err error
+		eachEntryFunc, err = sendToElasticSearchFunc(*esServers, *esIndex)
+		if err != nil {
+			fmt.Printf("create elastic search client to %s: %+v\n", *esServers, err)
+			return false
+		}
+	}
+
+	tailErr := pb.WithFilerClient(*tailFiler, grpcDialOption, func(client filer_pb.SeaweedFilerClient) error {
+
+		ctx, cancel := context.WithCancel(context.Background())
+		defer cancel()
+
+		stream, err := client.SubscribeMetadata(ctx, &filer_pb.SubscribeMetadataRequest{
+			ClientName: "tail",
+			PathPrefix: *tailTarget,
+			SinceNs:    time.Now().Add(-*tailStart).UnixNano(),
+		})
+		if err != nil {
+			return fmt.Errorf("listen: %v", err)
+		}
+
+		for {
+			resp, listenErr := stream.Recv()
+			if listenErr == io.EOF {
+				return nil
+			}
+			if listenErr != nil {
+				return listenErr
+			}
+			if !shouldPrint(resp) {
+				continue
+			}
+			if err = eachEntryFunc(resp); err != nil {
+				return err
+			}
+		}
+
+	})
+	if tailErr != nil {
+		fmt.Printf("tail %s: %v\n", *tailFiler, tailErr)
+	}
+
+	return true
+}
+
+type EsDocument struct {
+	Dir         string `json:"dir,omitempty"`
+	Name        string `json:"name,omitempty"`
+	IsDirectory bool   `json:"isDir,omitempty"`
+	Size        uint64 `json:"size,omitempty"`
+	Uid         uint32 `json:"uid,omitempty"`
+	Gid         uint32 `json:"gid,omitempty"`
+	UserName    string `json:"userName,omitempty"`
+	Collection  string `json:"collection,omitempty"`
+	Crtime      int64  `json:"crtime,omitempty"`
+	Mtime       int64  `json:"mtime,omitempty"`
+	Mime        string `json:"mime,omitempty"`
+}
+
+func toEsEntry(event *filer_pb.EventNotification) (*EsDocument, string) {
+	entry := event.NewEntry
+	dir, name := event.NewParentPath, entry.Name
+	id := util.Md5String([]byte(util.NewFullPath(dir, name)))
+	esEntry := &EsDocument{
+		Dir:         dir,
+		Name:        name,
+		IsDirectory: entry.IsDirectory,
+		Size:        entry.Attributes.FileSize,
+		Uid:         entry.Attributes.Uid,
+		Gid:         entry.Attributes.Gid,
+		UserName:    entry.Attributes.UserName,
+		Collection:  entry.Attributes.Collection,
+		Crtime:      entry.Attributes.Crtime,
+		Mtime:       entry.Attributes.Mtime,
+		Mime:        entry.Attributes.Mime,
+	}
+	return esEntry, id
+}
+
+func sendToElasticSearchFunc(servers string, esIndex string) (func(resp *filer_pb.SubscribeMetadataResponse) error, error) {
+	options := []elastic.ClientOptionFunc{}
+	options = append(options, elastic.SetURL(strings.Split(servers, ",")...))
+	options = append(options, elastic.SetSniff(false))
+	options = append(options, elastic.SetHealthcheck(false))
+	client, err := elastic.NewClient(options...)
+	if err != nil {
+		return nil, err
+	}
+	return func(resp *filer_pb.SubscribeMetadataResponse) error {
+		event := resp.EventNotification
+		if event.OldEntry != nil &&
+			(event.NewEntry == nil || resp.Directory != event.NewParentPath || event.OldEntry.Name != event.NewEntry.Name) {
+			// delete or not update the same file
+			dir, name := resp.Directory, event.OldEntry.Name
+			id := util.Md5String([]byte(util.NewFullPath(dir, name)))
+			println("delete", id)
+			_, err := client.Delete().Index(esIndex).Id(id).Do(context.Background())
+			return err
+		}
+		if event.NewEntry != nil {
+			// add a new file or update the same file
+			esEntry, id := toEsEntry(event)
+			value, err := jsoniter.Marshal(esEntry)
+			if err != nil {
+				return err
+			}
+			println(string(value))
+			_, err = client.Index().Index(esIndex).Id(id).BodyJson(string(value)).Do(context.Background())
+			return err
+		}
+		return nil
+	}, nil
+}
@@ -11,10 +11,10 @@ import (
 	_ "github.com/chrislusf/seaweedfs/weed/replication/sink/b2sink"
 	_ "github.com/chrislusf/seaweedfs/weed/replication/sink/filersink"
 	_ "github.com/chrislusf/seaweedfs/weed/replication/sink/gcssink"
+	_ "github.com/chrislusf/seaweedfs/weed/replication/sink/localsink"
 	_ "github.com/chrislusf/seaweedfs/weed/replication/sink/s3sink"
 	"github.com/chrislusf/seaweedfs/weed/replication/sub"
 	"github.com/chrislusf/seaweedfs/weed/util"
-	"github.com/spf13/viper"
 )

 func init() {
@@ -98,13 +98,19 @@ func runFilerReplicate(cmd *Command, args []string) bool {
 	replicator := replication.NewReplicator(config, "source.filer.", dataSink)

 	for {
-		key, m, err := notificationInput.ReceiveMessage()
+		key, m, onSuccessFn, onFailureFn, err := notificationInput.ReceiveMessage()
 		if err != nil {
 			glog.Errorf("receive %s: %+v", key, err)
+			if onFailureFn != nil {
+				onFailureFn()
+			}
 			continue
 		}
 		if key == "" {
 			// long poll received no messages
+			if onSuccessFn != nil {
+				onSuccessFn()
+			}
 			continue
 		}
 		if m.OldEntry != nil && m.NewEntry == nil {
@@ -116,14 +122,20 @@ func runFilerReplicate(cmd *Command, args []string) bool {
 		}
 		if err = replicator.Replicate(context.Background(), key, m); err != nil {
 			glog.Errorf("replicate %s: %+v", key, err)
+			if onFailureFn != nil {
+				onFailureFn()
+			}
 		} else {
 			glog.V(1).Infof("replicated %s", key)
+			if onSuccessFn != nil {
+				onSuccessFn()
+			}
 		}
 	}

 }

-func validateOneEnabledInput(config *viper.Viper) {
+func validateOneEnabledInput(config *util.ViperProxy) {
 	enabledInput := ""
 	for _, input := range sub.NotificationInputs {
 		if config.GetBool("notification." + input.GetName() + ".enabled") {
@@ -33,6 +33,8 @@ type SyncOptions struct {
 	bTtlSec         *int
 	aDebug          *bool
 	bDebug          *bool
+	aProxyByFiler   *bool
+	bProxyByFiler   *bool
 }

 var (
@@ -43,7 +45,7 @@ var (

 func init() {
 	cmdFilerSynchronize.Run = runFilerSynchronize // break init cycle
-	syncOptions.isActivePassive = cmdFilerSynchronize.Flag.Bool("isActivePassive", false, "one directional follow if true")
+	syncOptions.isActivePassive = cmdFilerSynchronize.Flag.Bool("isActivePassive", false, "one directional follow from A to B if true")
 	syncOptions.filerA = cmdFilerSynchronize.Flag.String("a", "", "filer A in one SeaweedFS cluster")
 	syncOptions.filerB = cmdFilerSynchronize.Flag.String("b", "", "filer B in the other SeaweedFS cluster")
 	syncOptions.aPath = cmdFilerSynchronize.Flag.String("a.path", "/", "directory to sync on filer A")
@@ -54,6 +56,8 @@ func init() {
 	syncOptions.bCollection = cmdFilerSynchronize.Flag.String("b.collection", "", "collection on filer B")
 	syncOptions.aTtlSec = cmdFilerSynchronize.Flag.Int("a.ttlSec", 0, "ttl in seconds on filer A")
 	syncOptions.bTtlSec = cmdFilerSynchronize.Flag.Int("b.ttlSec", 0, "ttl in seconds on filer B")
+	syncOptions.aProxyByFiler = cmdFilerSynchronize.Flag.Bool("a.filerProxy", false, "read and write file chunks by filer A instead of volume servers")
+	syncOptions.bProxyByFiler = cmdFilerSynchronize.Flag.Bool("b.filerProxy", false, "read and write file chunks by filer B instead of volume servers")
 	syncOptions.aDebug = cmdFilerSynchronize.Flag.Bool("a.debug", false, "debug mode to print out filer A received files")
 	syncOptions.bDebug = cmdFilerSynchronize.Flag.Bool("b.debug", false, "debug mode to print out filer B received files")
 	syncCpuProfile = cmdFilerSynchronize.Flag.String("cpuprofile", "", "cpu profile output file")
@@ -62,8 +66,8 @@ func init() {

 var cmdFilerSynchronize = &Command{
 	UsageLine: "filer.sync -a=<oneFilerHost>:<oneFilerPort> -b=<otherFilerHost>:<otherFilerPort>",
-	Short:     "continuously synchronize between two active-active or active-passive SeaweedFS clusters",
-	Long: `continuously synchronize file changes between two active-active or active-passive filers
+	Short:     "resumeable continuous synchronization between two active-active or active-passive SeaweedFS clusters",
+	Long: `resumeable continuous synchronization for file changes between two active-active or active-passive filers

 	filer.sync listens on filer notifications. If any file is updated, it will fetch the updated content,
 	and write to the other destination. Different from filer.replicate:
@@ -86,8 +90,9 @@ func runFilerSynchronize(cmd *Command, args []string) bool {

 	go func() {
 		for {
-			err := doSubscribeFilerMetaChanges(grpcDialOption, *syncOptions.filerA, *syncOptions.aPath, *syncOptions.filerB,
-				*syncOptions.bPath, *syncOptions.bReplication, *syncOptions.bCollection, *syncOptions.bTtlSec, *syncOptions.bDebug)
+			err := doSubscribeFilerMetaChanges(grpcDialOption, *syncOptions.filerA, *syncOptions.aPath, *syncOptions.aProxyByFiler,
+				*syncOptions.filerB, *syncOptions.bPath, *syncOptions.bReplication, *syncOptions.bCollection, *syncOptions.bTtlSec, *syncOptions.bProxyByFiler,
+				*syncOptions.bDebug)
 			if err != nil {
 				glog.Errorf("sync from %s to %s: %v", *syncOptions.filerA, *syncOptions.filerB, err)
 				time.Sleep(1747 * time.Millisecond)
@@ -98,8 +103,9 @@ func runFilerSynchronize(cmd *Command, args []string) bool {
 	if !*syncOptions.isActivePassive {
 		go func() {
 			for {
-				err := doSubscribeFilerMetaChanges(grpcDialOption, *syncOptions.filerB, *syncOptions.bPath, *syncOptions.filerA,
-					*syncOptions.aPath, *syncOptions.aReplication, *syncOptions.aCollection, *syncOptions.aTtlSec, *syncOptions.aDebug)
+				err := doSubscribeFilerMetaChanges(grpcDialOption, *syncOptions.filerB, *syncOptions.bPath, *syncOptions.bProxyByFiler,
+					*syncOptions.filerA, *syncOptions.aPath, *syncOptions.aReplication, *syncOptions.aCollection, *syncOptions.aTtlSec, *syncOptions.aProxyByFiler,
+					*syncOptions.aDebug)
 				if err != nil {
 					glog.Errorf("sync from %s to %s: %v", *syncOptions.filerB, *syncOptions.filerA, err)
 					time.Sleep(2147 * time.Millisecond)
@@ -113,8 +119,8 @@ func runFilerSynchronize(cmd *Command, args []string) bool {
 	return true
 }

-func doSubscribeFilerMetaChanges(grpcDialOption grpc.DialOption, sourceFiler, sourcePath, targetFiler, targetPath string,
-	replicationStr, collection string, ttlSec int, debug bool) error {
+func doSubscribeFilerMetaChanges(grpcDialOption grpc.DialOption, sourceFiler, sourcePath string, sourceReadChunkFromFiler bool, targetFiler, targetPath string,
+	replicationStr, collection string, ttlSec int, sinkWriteChunkByFiler, debug bool) error {

 	// read source filer signature
 	sourceFilerSignature, sourceErr := replication.ReadFilerSignature(grpcDialOption, sourceFiler)
@@ -138,9 +144,9 @@ func doSubscribeFilerMetaChanges(grpcDialOption grpc.DialOption, sourceFiler, so

 	// create filer sink
 	filerSource := &source.FilerSource{}
-	filerSource.DoInitialize(pb.ServerToGrpcAddress(sourceFiler), sourcePath)
+	filerSource.DoInitialize(sourceFiler, pb.ServerToGrpcAddress(sourceFiler), sourcePath, sourceReadChunkFromFiler)
 	filerSink := &filersink.FilerSink{}
-	filerSink.DoInitialize(pb.ServerToGrpcAddress(targetFiler), targetPath, replicationStr, collection, ttlSec, grpcDialOption)
+	filerSink.DoInitialize(targetFiler, pb.ServerToGrpcAddress(targetFiler), targetPath, replicationStr, collection, ttlSec, grpcDialOption, sinkWriteChunkByFiler)
 	filerSink.SetSourceFiler(filerSource)

 	processEventFn := func(resp *filer_pb.SubscribeMetadataResponse) error {
@@ -69,7 +69,7 @@ var cmdMaster = &Command{
 	Short:     "start a master server",
 	Long: `start a master server to provide volume=>location mapping service and sequence number of file ids

-	The configuration file "security.toml" is read from ".", "$HOME/.seaweedfs/", or "/etc/seaweedfs/", in that order.
+	The configuration file "security.toml" is read from ".", "$HOME/.seaweedfs/", "/usr/local/etc/seaweedfs/", or "/etc/seaweedfs/", in that order.

 	The example security.toml configuration file can be generated by "weed scaffold -config=security"

@@ -6,24 +6,24 @@ import (
 )

 type MountOptions struct {
-	filer                       *string
-	filerMountRootPath          *string
-	dir                         *string
-	dirAutoCreate               *bool
-	collection                  *string
-	replication                 *string
-	ttlSec                      *int
-	chunkSizeLimitMB            *int
-	concurrentWriters           *int
-	cacheDir                    *string
-	cacheSizeMB                 *int64
-	dataCenter                  *string
-	allowOthers                 *bool
-	umaskString                 *string
-	nonempty                    *bool
-	outsideContainerClusterMode *bool
-	uidMap                      *string
-	gidMap                      *string
+	filer              *string
+	filerMountRootPath *string
+	dir                *string
+	dirAutoCreate      *bool
+	collection         *string
+	replication        *string
+	ttlSec             *int
+	chunkSizeLimitMB   *int
+	concurrentWriters  *int
+	cacheDir           *string
+	cacheSizeMB        *int64
+	dataCenter         *string
+	allowOthers        *bool
+	umaskString        *string
+	nonempty           *bool
+	volumeServerAccess *string
+	uidMap             *string
+	gidMap             *string
 }

 var (
@@ -43,14 +43,14 @@ func init() {
 	mountOptions.replication = cmdMount.Flag.String("replication", "", "replication(e.g. 000, 001) to create to files. If empty, let filer decide.")
 	mountOptions.ttlSec = cmdMount.Flag.Int("ttl", 0, "file ttl in seconds")
 	mountOptions.chunkSizeLimitMB = cmdMount.Flag.Int("chunkSizeLimitMB", 2, "local write buffer size, also chunk large files")
-	mountOptions.concurrentWriters = cmdMount.Flag.Int("concurrentWriters", 0, "limit concurrent goroutine writers if not 0")
+	mountOptions.concurrentWriters = cmdMount.Flag.Int("concurrentWriters", 128, "limit concurrent goroutine writers if not 0")
 	mountOptions.cacheDir = cmdMount.Flag.String("cacheDir", os.TempDir(), "local cache directory for file chunks and meta data")
 	mountOptions.cacheSizeMB = cmdMount.Flag.Int64("cacheCapacityMB", 1000, "local file chunk cache capacity in MB (0 will disable cache)")
 	mountOptions.dataCenter = cmdMount.Flag.String("dataCenter", "", "prefer to write to the data center")
 	mountOptions.allowOthers = cmdMount.Flag.Bool("allowOthers", true, "allows other users to access the file system")
 	mountOptions.umaskString = cmdMount.Flag.String("umask", "022", "octal umask, e.g., 022, 0111")
 	mountOptions.nonempty = cmdMount.Flag.Bool("nonempty", false, "allows the mounting over a non-empty directory")
-	mountOptions.outsideContainerClusterMode = cmdMount.Flag.Bool("outsideContainerClusterMode", false, "allows other users to access volume servers with publicUrl")
+	mountOptions.volumeServerAccess = cmdMount.Flag.String("volumeServerAccess", "direct", "access volume servers by [direct|publicUrl|filerProxy]")
 	mountOptions.uidMap = cmdMount.Flag.String("map.uid", "", "map local uid to uid on filer, comma-separated <local_uid>:<filer_uid>")
 	mountOptions.gidMap = cmdMount.Flag.String("map.gid", "", "map local gid to gid on filer, comma-separated <local_gid>:<filer_gid>")

@@ -58,6 +58,7 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 		return true
 	}

+	util.LoadConfiguration("security", false)
 	// try to connect to filer, filerBucketsPath may be useful later
 	grpcDialOption := security.LoadClientTLS(util.GetViper(), "grpc.client")
 	var cipher bool
@@ -78,8 +79,6 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 	dir := util.ResolvePath(*option.dir)
 	chunkSizeLimitMB := *mountOptions.chunkSizeLimitMB

-	util.LoadConfiguration("security", false)
-
 	fmt.Printf("This is SeaweedFS version %s %s %s\n", util.Version(), runtime.GOOS, runtime.GOARCH)
 	if dir == "" {
 		fmt.Printf("Please specify the mount directory via \"-dir\"")
@@ -101,9 +100,9 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 	uid, gid := uint32(0), uint32(0)
 	mountMode := os.ModeDir | 0777
 	if err == nil {
-		mountMode = os.ModeDir | fileInfo.Mode()
+		mountMode = os.ModeDir | os.FileMode(0777)&^umask
 		uid, gid = util.GetFileUidGid(fileInfo)
-		fmt.Printf("mount point owner uid=%d gid=%d mode=%s\n", uid, gid, fileInfo.Mode())
+		fmt.Printf("mount point owner uid=%d gid=%d mode=%s\n", uid, gid, mountMode)
 	} else {
 		fmt.Printf("can not stat %s\n", dir)
 		return false
@@ -151,6 +150,8 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 		fuse.MaxReadahead(1024 * 128),
 		fuse.AsyncRead(),
 		fuse.WritebackCache(),
+		fuse.MaxBackground(128),
+		fuse.CongestionThreshold(128),
 	}

 	options = append(options, osSpecificMountOptions()...)
@@ -168,27 +169,29 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 	}

 	seaweedFileSystem := filesys.NewSeaweedFileSystem(&filesys.Option{
-		FilerGrpcAddress:            filerGrpcAddress,
-		GrpcDialOption:              grpcDialOption,
-		FilerMountRootPath:          mountRoot,
-		Collection:                  *option.collection,
-		Replication:                 *option.replication,
-		TtlSec:                      int32(*option.ttlSec),
-		ChunkSizeLimit:              int64(chunkSizeLimitMB) * 1024 * 1024,
-		ConcurrentWriters:           *option.concurrentWriters,
-		CacheDir:                    *option.cacheDir,
-		CacheSizeMB:                 *option.cacheSizeMB,
-		DataCenter:                  *option.dataCenter,
-		EntryCacheTtl:               3 * time.Second,
-		MountUid:                    uid,
-		MountGid:                    gid,
-		MountMode:                   mountMode,
-		MountCtime:                  fileInfo.ModTime(),
-		MountMtime:                  time.Now(),
-		Umask:                       umask,
-		OutsideContainerClusterMode: *mountOptions.outsideContainerClusterMode,
-		Cipher:                      cipher,
-		UidGidMapper:                uidGidMapper,
+		MountDirectory:     dir,
+		FilerAddress:       filer,
+		FilerGrpcAddress:   filerGrpcAddress,
+		GrpcDialOption:     grpcDialOption,
+		FilerMountRootPath: mountRoot,
+		Collection:         *option.collection,
+		Replication:        *option.replication,
+		TtlSec:             int32(*option.ttlSec),
+		ChunkSizeLimit:     int64(chunkSizeLimitMB) * 1024 * 1024,
+		ConcurrentWriters:  *option.concurrentWriters,
+		CacheDir:           *option.cacheDir,
+		CacheSizeMB:        *option.cacheSizeMB,
+		DataCenter:         *option.dataCenter,
+		EntryCacheTtl:      3 * time.Second,
+		MountUid:           uid,
+		MountGid:           gid,
+		MountMode:          mountMode,
+		MountCtime:         fileInfo.ModTime(),
+		MountMtime:         time.Now(),
+		Umask:              umask,
+		VolumeServerAccess: *mountOptions.volumeServerAccess,
+		Cipher:             cipher,
+		UidGidMapper:       uidGidMapper,
 	})

 	// mount
@@ -205,7 +208,9 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 	})

 	glog.V(0).Infof("mounted %s%s to %s", filer, mountRoot, dir)
-	err = fs.Serve(c, seaweedFileSystem)
+	server := fs.New(c, nil)
+	seaweedFileSystem.Server = server
+	err = server.Serve(seaweedFileSystem)

 	// check if the mount process has an error to report
 	<-c.Ready
@@ -23,13 +23,14 @@ var (
 )

 type S3Options struct {
-	filer           *string
-	port            *int
-	config          *string
-	domainName      *string
-	tlsPrivateKey   *string
-	tlsCertificate  *string
-	metricsHttpPort *int
+	filer            *string
+	port             *int
+	config           *string
+	domainName       *string
+	tlsPrivateKey    *string
+	tlsCertificate   *string
+	metricsHttpPort  *int
+	allowEmptyFolder *bool
 }

 func init() {
@@ -41,6 +42,7 @@ func init() {
 	s3StandaloneOptions.tlsPrivateKey = cmdS3.Flag.String("key.file", "", "path to the TLS private key file")
 	s3StandaloneOptions.tlsCertificate = cmdS3.Flag.String("cert.file", "", "path to the TLS certificate file")
 	s3StandaloneOptions.metricsHttpPort = cmdS3.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
+	s3StandaloneOptions.allowEmptyFolder = cmdS3.Flag.Bool("allowEmptyFolder", false, "allow empty folders")
 }

 var cmdS3 = &Command{
@@ -181,6 +183,7 @@ func (s3opt *S3Options) startS3Server() bool {
 		DomainName:       *s3opt.domainName,
 		BucketsPath:      filerBucketsPath,
 		GrpcDialOption:   grpcDialOption,
+		AllowEmptyFolder: *s3opt.allowEmptyFolder,
 	})
 	if s3ApiServer_err != nil {
 		glog.Fatalf("S3 API Server startup error: %v", s3ApiServer_err)
@@ -44,6 +44,8 @@ func runScaffold(cmd *Command, args []string) bool {
 		content = SECURITY_TOML_EXAMPLE
 	case "master":
 		content = MASTER_TOML_EXAMPLE
+	case "shell":
+		content = SHELL_TOML_EXAMPLE
 	}
 	if content == "" {
 		println("need a valid -config option")
@@ -85,9 +87,21 @@ buckets_folder = "/buckets"
 # local on disk, mostly for simple single-machine setup, fairly scalable
 # faster than previous leveldb, recommended.
 enabled = true
-dir = "."					# directory to store level db files
+dir = "./filerldb2"					# directory to store level db files

-[mysql]  # or tidb
+[leveldb3]
+# similar to leveldb2.
+# each bucket has its own meta store.
+enabled = false
+dir = "./filerldb3"					# directory to store level db files
+
+[rocksdb]
+# local on disk, similar to leveldb
+# since it is using a C wrapper, you need to install rocksdb and build it by yourself
+enabled = false
+dir = "./filerrdb"					# directory to store rocksdb files
+
+[mysql]  # or memsql, tidb
 # CREATE TABLE IF NOT EXISTS filemeta (
 #   dirhash     BIGINT         COMMENT 'first 64 bits of MD5 hash value of directory field',
 #   name        VARCHAR(1000)  COMMENT 'directory or file name',
@@ -104,9 +118,31 @@ password = ""
 database = ""              # create or use an existing database
 connection_max_idle = 2
 connection_max_open = 100
+connection_max_lifetime_seconds = 0
 interpolateParams = false

-[postgres] # or cockroachdb
+[mysql2]  # or memsql, tidb
+enabled = false
+createTable = """
+  CREATE TABLE IF NOT EXISTS %s (
+    dirhash BIGINT, 
+    name VARCHAR(1000), 
+    directory TEXT, 
+    meta LONGBLOB, 
+    PRIMARY KEY (dirhash, name)
+  ) DEFAULT CHARSET=utf8;
+"""
+hostname = "localhost"
+port = 3306
+username = "root"
+password = ""
+database = ""              # create or use an existing database
+connection_max_idle = 2
+connection_max_open = 100
+connection_max_lifetime_seconds = 0
+interpolateParams = false
+
+[postgres] # or cockroachdb, YugabyteDB
 # CREATE TABLE IF NOT EXISTS filemeta (
 #   dirhash     BIGINT,
 #   name        VARCHAR(65535),
@@ -119,7 +155,29 @@ hostname = "localhost"
 port = 5432
 username = "postgres"
 password = ""
-database = ""              # create or use an existing database
+database = "postgres"          # create or use an existing database
+schema = ""
+sslmode = "disable"
+connection_max_idle = 100
+connection_max_open = 100
+
+[postgres2]
+enabled = false
+createTable = """
+  CREATE TABLE IF NOT EXISTS %s (
+    dirhash   BIGINT, 
+    name      VARCHAR(65535), 
+    directory VARCHAR(65535), 
+    meta      bytea, 
+    PRIMARY KEY (dirhash, name)
+  );
+"""
+hostname = "localhost"
+port = 5432
+username = "postgres"
+password = ""
+database = "postgres"          # create or use an existing database
+schema = ""
 sslmode = "disable"
 connection_max_idle = 100
 connection_max_open = 100
@@ -138,12 +196,21 @@ hosts=[
 ]
 username=""
 password=""
+# This changes the data layout. Only add new directories. Removing/Updating will cause data loss.
+superLargeDirectories = []
+
+[hbase]
+enabled = false
+zkquorum = ""
+table = "seaweedfs"

 [redis2]
 enabled = false
 address  = "localhost:6379"
 password = ""
 database = 0
+# This changes the data layout. Only add new directories. Removing/Updating will cause data loss.
+superLargeDirectories = []

 [redis_cluster2]
 enabled = false
@@ -157,9 +224,11 @@ addresses = [
 ]
 password = ""
 # allows reads from slave servers or the master, but all writes still go to the master
-readOnly = true
+readOnly = false
 # automatically use the closest Redis server for reads
-routeByLatency = true
+routeByLatency = false
+# This changes the data layout. Only add new directories. Removing/Updating will cause data loss.
+superLargeDirectories = []

 [etcd]
 enabled = false
@@ -185,6 +254,28 @@ sniff_enabled = false
 healthcheck_enabled = false
 # increase the value is recommend, be sure the value in Elastic is greater or equal here
 index.max_result_window = 10000
+
+
+
+##########################
+##########################
+# To add path-specific filer store:
+#
+# 1. Add a name following the store type separated by a dot ".". E.g., cassandra.tmp
+# 2. Add a location configuraiton. E.g., location = "/tmp/"
+# 3. Copy and customize all other configurations. 
+#     Make sure they are not the same if using the same store type!
+# 4. Set enabled to true
+#
+# The following is just using cassandra as an example
+##########################
+[redis2.tmp]
+enabled = false
+location = "/tmp/"
+address  = "localhost:6379"
+password = ""
+database = 1
+
 `

 	NOTIFICATION_TOML_EXAMPLE = `
@@ -238,7 +329,8 @@ enabled = false
 # This URL will Dial the RabbitMQ server at the URL in the environment
 # variable RABBIT_SERVER_URL and open the exchange "myexchange".
 # The exchange must have already been created by some other means, like
-# the RabbitMQ management plugin.
+# the RabbitMQ management plugin. Сreate myexchange of type fanout and myqueue then
+# create binding myexchange => myqueue
 topic_url = "rabbit://myexchange"
 sub_url = "rabbit://myqueue"
 `
@@ -259,6 +351,16 @@ grpcAddress = "localhost:18888"
 # i.e., all files with this "prefix" are sent to notification message queue.
 directory = "/buckets"

+[sink.local]
+enabled = false
+directory = "/data"
+
+[sink.local_incremental]
+# all replicated files are under modified time as yyyy-mm-dd directories
+# so each date directory contains all new and updated files.
+enabled = false
+directory = "/backup"
+
 [sink.filer]
 enabled = false
 grpcAddress = "localhost:18888"
@@ -426,5 +528,19 @@ copy_other = 1            # create n x 1 = n actual volumes
 # if you are doing your own replication or periodic sync of volumes.
 treat_replication_as_minimums = false

+`
+	SHELL_TOML_EXAMPLE = `
+
+[cluster]
+default = "c1"
+
+[cluster.c1]
+master = "localhost:9333"    # comma-separated master servers
+filer = "localhost:8888"     # filer host and port
+
+[cluster.c2]
+master = ""
+filer = ""
+
 `
 )
@@ -61,6 +61,7 @@ var (
 	serverMetricsHttpPort     = cmdServer.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")

 	// pulseSeconds              = cmdServer.Flag.Int("pulseSeconds", 5, "number of seconds between heartbeats")
+	isStartingMasterServer = cmdServer.Flag.Bool("master", true, "whether to start master server")
 	isStartingVolumeServer = cmdServer.Flag.Bool("volume", true, "whether to start volume server")
 	isStartingFiler        = cmdServer.Flag.Bool("filer", false, "whether to start filer")
 	isStartingS3           = cmdServer.Flag.Bool("s3", false, "whether to start S3 gateway")
@@ -94,6 +95,7 @@ func init() {
 	filerOptions.dirListingLimit = cmdServer.Flag.Int("filer.dirListLimit", 1000, "limit sub dir listing size")
 	filerOptions.cipher = cmdServer.Flag.Bool("filer.encryptVolumeData", false, "encrypt data on volume servers")
 	filerOptions.peers = cmdServer.Flag.String("filer.peers", "", "all filers sharing the same filer store in comma separated ip:port list")
+	filerOptions.saveToFilerLimit = cmdServer.Flag.Int("filer.saveToFilerLimit", 0, "Small files smaller than this limit can be cached in filer store.")

 	serverOptions.v.port = cmdServer.Flag.Int("volume.port", 8080, "volume server http listen port")
 	serverOptions.v.publicPort = cmdServer.Flag.Int("volume.port.public", 0, "volume server public port")
@@ -105,12 +107,14 @@ func init() {
 	serverOptions.v.publicUrl = cmdServer.Flag.String("volume.publicUrl", "", "publicly accessible address")
 	serverOptions.v.preStopSeconds = cmdServer.Flag.Int("volume.preStopSeconds", 10, "number of seconds between stop send heartbeats and stop volume server")
 	serverOptions.v.pprof = cmdServer.Flag.Bool("volume.pprof", false, "enable pprof http handlers. precludes --memprofile and --cpuprofile")
+	serverOptions.v.idxFolder = cmdServer.Flag.String("volume.dir.idx", "", "directory to store .idx files")

 	s3Options.port = cmdServer.Flag.Int("s3.port", 8333, "s3 server http listen port")
 	s3Options.domainName = cmdServer.Flag.String("s3.domainName", "", "suffix of the host name in comma separated list, {bucket}.{domainName}")
 	s3Options.tlsPrivateKey = cmdServer.Flag.String("s3.key.file", "", "path to the TLS private key file")
 	s3Options.tlsCertificate = cmdServer.Flag.String("s3.cert.file", "", "path to the TLS certificate file")
 	s3Options.config = cmdServer.Flag.String("s3.config", "", "path to the config file")
+	s3Options.allowEmptyFolder = cmdServer.Flag.Bool("s3.allowEmptyFolder", false, "allow empty folders")

 	msgBrokerOptions.port = cmdServer.Flag.Int("msgBroker.port", 17777, "broker gRPC listen port")

@@ -221,7 +225,11 @@ func runServer(cmd *Command, args []string) bool {

 	}

-	startMaster(masterOptions, serverWhiteList)
+	if *isStartingMasterServer {
+		go startMaster(masterOptions, serverWhiteList)
+	}
+
+	select {}

 	return true
 }
@@ -11,12 +11,14 @@ import (
 var (
 	shellOptions      shell.ShellOptions
 	shellInitialFiler *string
+	shellCluster      *string
 )

 func init() {
 	cmdShell.Run = runShell // break init cycle
-	shellOptions.Masters = cmdShell.Flag.String("master", "localhost:9333", "comma-separated master servers")
-	shellInitialFiler = cmdShell.Flag.String("filer", "localhost:8888", "filer host and port")
+	shellOptions.Masters = cmdShell.Flag.String("master", "", "comma-separated master servers, e.g. localhost:9333")
+	shellInitialFiler = cmdShell.Flag.String("filer", "", "filer host and port, e.g. localhost:8888")
+	shellCluster = cmdShell.Flag.String("cluster", "", "cluster defined in shell.toml")
 }

 var cmdShell = &Command{
@@ -24,6 +26,8 @@ var cmdShell = &Command{
 	Short:     "run interactive administrative commands",
 	Long: `run interactive administrative commands.

+	Generate shell.toml via "weed scaffold -config=shell"
+
  `,
 }

@@ -32,6 +36,23 @@ func runShell(command *Command, args []string) bool {
 	util.LoadConfiguration("security", false)
 	shellOptions.GrpcDialOption = security.LoadClientTLS(util.GetViper(), "grpc.client")

+	if *shellOptions.Masters == "" && *shellInitialFiler == "" {
+		util.LoadConfiguration("shell", false)
+		v := util.GetViper()
+		cluster := v.GetString("cluster.default")
+		if *shellCluster != "" {
+			cluster = *shellCluster
+		}
+		if cluster == "" {
+			*shellOptions.Masters, *shellInitialFiler = "localhost:9333", "localhost:8888"
+		} else {
+			*shellOptions.Masters = v.GetString("cluster." + cluster + ".master")
+			*shellInitialFiler = v.GetString("cluster." + cluster + ".filer")
+		}
+	}
+
+	fmt.Printf("master: %s filer: %s\n", *shellOptions.Masters, *shellInitialFiler)
+
 	var err error
 	shellOptions.FilerHost, shellOptions.FilerPort, err = util.ParseHostPort(*shellInitialFiler)
 	if err != nil {
@@ -1,8 +1,12 @@
 package command

 import (
+	"context"
 	"encoding/json"
 	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/master_pb"
+	"google.golang.org/grpc"
 	"os"
 	"path/filepath"

@@ -65,6 +69,15 @@ func runUpload(cmd *Command, args []string) bool {
 	util.LoadConfiguration("security", false)
 	grpcDialOption := security.LoadClientTLS(util.GetViper(), "grpc.client")

+	defaultCollection, err := readMasterConfiguration(grpcDialOption, *upload.master)
+	if err != nil {
+		fmt.Printf("upload: %v", err)
+		return false
+	}
+	if *upload.replication == "" {
+		*upload.replication = defaultCollection
+	}
+
 	if len(args) == 0 {
 		if *upload.dir == "" {
 			return false
@@ -104,3 +117,15 @@ func runUpload(cmd *Command, args []string) bool {
 	}
 	return true
 }
+
+func readMasterConfiguration(grpcDialOption grpc.DialOption, masterAddress string) (replication string, err error) {
+	err = pb.WithMasterClient(masterAddress, grpcDialOption, func(client master_pb.SeaweedClient) error {
+		resp, err := client.GetMasterConfiguration(context.Background(), &master_pb.GetMasterConfigurationRequest{})
+		if err != nil {
+			return fmt.Errorf("get master %s configuration: %v", masterAddress, err)
+		}
+		replication = resp.DefaultReplication
+		return nil
+	})
+	return
+}
@@ -39,6 +39,7 @@ type VolumeServerOptions struct {
 	publicPort            *int
 	folders               []string
 	folderMaxLimits       []int
+	idxFolder             *string
 	ip                    *string
 	publicUrl             *string
 	bindIp                *string
@@ -83,6 +84,7 @@ func init() {
 	v.fileSizeLimitMB = cmdVolume.Flag.Int("fileSizeLimitMB", 256, "limit file size to avoid out of memory")
 	v.pprof = cmdVolume.Flag.Bool("pprof", false, "enable pprof http handlers. precludes --memprofile and --cpuprofile")
 	v.metricsHttpPort = cmdVolume.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
+	v.idxFolder = cmdVolume.Flag.String("dir.idx", "", "directory to store .idx files")
 }

 var cmdVolume = &Command{
@@ -211,6 +213,7 @@ func (v VolumeServerOptions) startVolumeServer(volumeFolders, maxVolumeCounts, v
 	volumeServer := weed_server.NewVolumeServer(volumeMux, publicVolumeMux,
 		*v.ip, *v.port, *v.publicUrl,
 		v.folders, v.folderMaxLimits, v.minFreeSpacePercents,
+		*v.idxFolder,
 		volumeNeedleMapKind,
 		strings.Split(masters, ","), 5, *v.dataCenter, *v.rack,
 		v.whiteList,
@@ -1,113 +0,0 @@
-package command
-
-import (
-	"context"
-	"fmt"
-	"io"
-	"path/filepath"
-	"strings"
-	"time"
-
-	"github.com/chrislusf/seaweedfs/weed/pb"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"github.com/chrislusf/seaweedfs/weed/security"
-	"github.com/chrislusf/seaweedfs/weed/util"
-)
-
-func init() {
-	cmdWatch.Run = runWatch // break init cycle
-}
-
-var cmdWatch = &Command{
-	UsageLine: "watch [-filer=localhost:8888] [-target=/]",
-	Short:     "see recent changes on a filer",
-	Long: `See recent changes on a filer.
-
-  `,
-}
-
-var (
-	watchFiler   = cmdWatch.Flag.String("filer", "localhost:8888", "filer hostname:port")
-	watchTarget  = cmdWatch.Flag.String("pathPrefix", "/", "path to a folder or file, or common prefix for the folders or files on filer")
-	watchStart   = cmdWatch.Flag.Duration("timeAgo", 0, "start time before now. \"300ms\", \"1.5h\" or \"2h45m\". Valid time units are \"ns\", \"us\" (or \"µs\"), \"ms\", \"s\", \"m\", \"h\"")
-	watchPattern = cmdWatch.Flag.String("pattern", "", "full path or just filename pattern, ex: \"/home/?opher\", \"*.pdf\", see https://golang.org/pkg/path/filepath/#Match ")
-)
-
-func runWatch(cmd *Command, args []string) bool {
-
-	grpcDialOption := security.LoadClientTLS(util.GetViper(), "grpc.client")
-
-	var filterFunc func(dir, fname string) bool
-	if *watchPattern != "" {
-		if strings.Contains(*watchPattern, "/") {
-			println("watch path pattern", *watchPattern)
-			filterFunc = func(dir, fname string) bool {
-				matched, err := filepath.Match(*watchPattern, dir+"/"+fname)
-				if err != nil {
-					fmt.Printf("error: %v", err)
-				}
-				return matched
-			}
-		} else {
-			println("watch file pattern", *watchPattern)
-			filterFunc = func(dir, fname string) bool {
-				matched, err := filepath.Match(*watchPattern, fname)
-				if err != nil {
-					fmt.Printf("error: %v", err)
-				}
-				return matched
-			}
-		}
-	}
-
-	shouldPrint := func(resp *filer_pb.SubscribeMetadataResponse) bool {
-		if filterFunc == nil {
-			return true
-		}
-		if resp.EventNotification.OldEntry == nil && resp.EventNotification.NewEntry == nil {
-			return false
-		}
-		if resp.EventNotification.OldEntry != nil && filterFunc(resp.Directory, resp.EventNotification.OldEntry.Name) {
-			return true
-		}
-		if resp.EventNotification.NewEntry != nil && filterFunc(resp.EventNotification.NewParentPath, resp.EventNotification.NewEntry.Name) {
-			return true
-		}
-		return false
-	}
-
-	watchErr := pb.WithFilerClient(*watchFiler, grpcDialOption, func(client filer_pb.SeaweedFilerClient) error {
-
-		ctx, cancel := context.WithCancel(context.Background())
-		defer cancel()
-
-		stream, err := client.SubscribeMetadata(ctx, &filer_pb.SubscribeMetadataRequest{
-			ClientName: "watch",
-			PathPrefix: *watchTarget,
-			SinceNs:    time.Now().Add(-*watchStart).UnixNano(),
-		})
-		if err != nil {
-			return fmt.Errorf("listen: %v", err)
-		}
-
-		for {
-			resp, listenErr := stream.Recv()
-			if listenErr == io.EOF {
-				return nil
-			}
-			if listenErr != nil {
-				return listenErr
-			}
-			if !shouldPrint(resp) {
-				continue
-			}
-			fmt.Printf("dir:%s %+v\n", resp.Directory, resp.EventNotification)
-		}
-
-	})
-	if watchErr != nil {
-		fmt.Printf("watch %s: %v\n", *watchFiler, watchErr)
-	}
-
-	return true
-}
@@ -9,19 +9,33 @@ import (
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
 	"github.com/chrislusf/seaweedfs/weed/util"
 	"strings"
+	"sync"
 )

-type AbstractSqlStore struct {
-	DB                      *sql.DB
-	SqlInsert               string
-	SqlUpdate               string
-	SqlFind                 string
-	SqlDelete               string
-	SqlDeleteFolderChildren string
-	SqlListExclusive        string
-	SqlListInclusive        string
+type SqlGenerator interface {
+	GetSqlInsert(bucket string) string
+	GetSqlUpdate(bucket string) string
+	GetSqlFind(bucket string) string
+	GetSqlDelete(bucket string) string
+	GetSqlDeleteFolderChildren(bucket string) string
+	GetSqlListExclusive(bucket string) string
+	GetSqlListInclusive(bucket string) string
+	GetSqlCreateTable(bucket string) string
+	GetSqlDropTable(bucket string) string
 }

+type AbstractSqlStore struct {
+	SqlGenerator
+	DB                 *sql.DB
+	SupportBucketTable bool
+	dbs                map[string]bool
+	dbsLock            sync.Mutex
+}
+
+const (
+	DEFAULT_TABLE = "filemeta"
+)
+
 type TxOrDB interface {
 	ExecContext(ctx context.Context, query string, args ...interface{}) (sql.Result, error)
 	QueryRowContext(ctx context.Context, query string, args ...interface{}) *sql.Row
@@ -52,16 +66,65 @@ func (store *AbstractSqlStore) RollbackTransaction(ctx context.Context) error {
 	return nil
 }

-func (store *AbstractSqlStore) getTxOrDB(ctx context.Context) TxOrDB {
+func (store *AbstractSqlStore) getTxOrDB(ctx context.Context, fullpath util.FullPath, isForChildren bool) (txOrDB TxOrDB, bucket string, shortPath util.FullPath, err error) {
+
+	shortPath = fullpath
+	bucket = DEFAULT_TABLE
+
 	if tx, ok := ctx.Value("tx").(*sql.Tx); ok {
-		return tx
+		txOrDB = tx
+	} else {
+		txOrDB = store.DB
 	}
-	return store.DB
+
+	if !store.SupportBucketTable {
+		return
+	}
+
+	if !strings.HasPrefix(string(fullpath), "/buckets/") {
+		return
+	}
+
+	// detect bucket
+	bucketAndObjectKey := string(fullpath)[len("/buckets/"):]
+	t := strings.Index(bucketAndObjectKey, "/")
+	if t < 0 && !isForChildren {
+		return
+	}
+	bucket = bucketAndObjectKey
+	shortPath = "/"
+	if t > 0 {
+		bucket = bucketAndObjectKey[:t]
+		shortPath = util.FullPath(bucketAndObjectKey[t:])
+	}
+
+	if isValidBucket(bucket) {
+		store.dbsLock.Lock()
+		defer store.dbsLock.Unlock()
+
+		if store.dbs == nil {
+			store.dbs = make(map[string]bool)
+		}
+
+		if _, found := store.dbs[bucket]; !found {
+			if err = store.CreateTable(ctx, bucket); err != nil {
+				store.dbs[bucket] = true
+			}
+		}
+
+	}
+
+	return
 }

 func (store *AbstractSqlStore) InsertEntry(ctx context.Context, entry *filer.Entry) (err error) {

-	dir, name := entry.FullPath.DirAndName()
+	db, bucket, shortPath, err := store.getTxOrDB(ctx, entry.FullPath, false)
+	if err != nil {
+		return fmt.Errorf("findDB %s : %v", entry.FullPath, err)
+	}
+
+	dir, name := shortPath.DirAndName()
 	meta, err := entry.EncodeAttributesAndChunks()
 	if err != nil {
 		return fmt.Errorf("encode %s: %s", entry.FullPath, err)
@@ -71,19 +134,20 @@ func (store *AbstractSqlStore) InsertEntry(ctx context.Context, entry *filer.Ent
 		meta = util.MaybeGzipData(meta)
 	}

-	res, err := store.getTxOrDB(ctx).ExecContext(ctx, store.SqlInsert, util.HashStringToLong(dir), name, dir, meta)
+	res, err := db.ExecContext(ctx, store.GetSqlInsert(bucket), util.HashStringToLong(dir), name, dir, meta)
 	if err == nil {
 		return
 	}

 	if !strings.Contains(strings.ToLower(err.Error()), "duplicate") {
-		return fmt.Errorf("kv insert: %s", err)
+		// return fmt.Errorf("insert: %s", err)
+		// skip this since the error can be in a different language
 	}

 	// now the insert failed possibly due to duplication constraints
 	glog.V(1).Infof("insert %s falls back to update: %v", entry.FullPath, err)

-	res, err = store.getTxOrDB(ctx).ExecContext(ctx, store.SqlUpdate, meta, util.HashStringToLong(dir), name, dir)
+	res, err = db.ExecContext(ctx, store.GetSqlUpdate(bucket), meta, util.HashStringToLong(dir), name, dir)
 	if err != nil {
 		return fmt.Errorf("upsert %s: %s", entry.FullPath, err)
 	}
@@ -98,13 +162,18 @@ func (store *AbstractSqlStore) InsertEntry(ctx context.Context, entry *filer.Ent

 func (store *AbstractSqlStore) UpdateEntry(ctx context.Context, entry *filer.Entry) (err error) {

-	dir, name := entry.FullPath.DirAndName()
+	db, bucket, shortPath, err := store.getTxOrDB(ctx, entry.FullPath, false)
+	if err != nil {
+		return fmt.Errorf("findDB %s : %v", entry.FullPath, err)
+	}
+
+	dir, name := shortPath.DirAndName()
 	meta, err := entry.EncodeAttributesAndChunks()
 	if err != nil {
 		return fmt.Errorf("encode %s: %s", entry.FullPath, err)
 	}

-	res, err := store.getTxOrDB(ctx).ExecContext(ctx, store.SqlUpdate, meta, util.HashStringToLong(dir), name, dir)
+	res, err := db.ExecContext(ctx, store.GetSqlUpdate(bucket), meta, util.HashStringToLong(dir), name, dir)
 	if err != nil {
 		return fmt.Errorf("update %s: %s", entry.FullPath, err)
 	}
@@ -118,8 +187,13 @@ func (store *AbstractSqlStore) UpdateEntry(ctx context.Context, entry *filer.Ent

 func (store *AbstractSqlStore) FindEntry(ctx context.Context, fullpath util.FullPath) (*filer.Entry, error) {

-	dir, name := fullpath.DirAndName()
-	row := store.getTxOrDB(ctx).QueryRowContext(ctx, store.SqlFind, util.HashStringToLong(dir), name, dir)
+	db, bucket, shortPath, err := store.getTxOrDB(ctx, fullpath, false)
+	if err != nil {
+		return nil, fmt.Errorf("findDB %s : %v", fullpath, err)
+	}
+
+	dir, name := shortPath.DirAndName()
+	row := db.QueryRowContext(ctx, store.GetSqlFind(bucket), util.HashStringToLong(dir), name, dir)

 	var data []byte
 	if err := row.Scan(&data); err != nil {
@@ -141,9 +215,14 @@ func (store *AbstractSqlStore) FindEntry(ctx context.Context, fullpath util.Full

 func (store *AbstractSqlStore) DeleteEntry(ctx context.Context, fullpath util.FullPath) error {

-	dir, name := fullpath.DirAndName()
+	db, bucket, shortPath, err := store.getTxOrDB(ctx, fullpath, false)
+	if err != nil {
+		return fmt.Errorf("findDB %s : %v", fullpath, err)
+	}

-	res, err := store.getTxOrDB(ctx).ExecContext(ctx, store.SqlDelete, util.HashStringToLong(dir), name, dir)
+	dir, name := shortPath.DirAndName()
+
+	res, err := db.ExecContext(ctx, store.GetSqlDelete(bucket), util.HashStringToLong(dir), name, dir)
 	if err != nil {
 		return fmt.Errorf("delete %s: %s", fullpath, err)
 	}
@@ -158,7 +237,23 @@ func (store *AbstractSqlStore) DeleteEntry(ctx context.Context, fullpath util.Fu

 func (store *AbstractSqlStore) DeleteFolderChildren(ctx context.Context, fullpath util.FullPath) error {

-	res, err := store.getTxOrDB(ctx).ExecContext(ctx, store.SqlDeleteFolderChildren, util.HashStringToLong(string(fullpath)), fullpath)
+	db, bucket, shortPath, err := store.getTxOrDB(ctx, fullpath, true)
+	if err != nil {
+		return fmt.Errorf("findDB %s : %v", fullpath, err)
+	}
+
+	if isValidBucket(bucket) && shortPath == "/" {
+		if err = store.deleteTable(ctx, bucket); err == nil {
+			store.dbsLock.Lock()
+			delete(store.dbs, bucket)
+			store.dbsLock.Unlock()
+			return nil
+		} else {
+			return err
+		}
+	}
+
+	res, err := db.ExecContext(ctx, store.GetSqlDeleteFolderChildren(bucket), util.HashStringToLong(string(shortPath)), fullpath)
 	if err != nil {
 		return fmt.Errorf("deleteFolderChildren %s: %s", fullpath, err)
 	}
@@ -171,15 +266,21 @@ func (store *AbstractSqlStore) DeleteFolderChildren(ctx context.Context, fullpat
 	return nil
 }

-func (store *AbstractSqlStore) ListDirectoryPrefixedEntries(ctx context.Context, fullpath util.FullPath, startFileName string, inclusive bool, limit int, prefix string) (entries []*filer.Entry, err error) {
-	sqlText := store.SqlListExclusive
-	if inclusive {
-		sqlText = store.SqlListInclusive
+func (store *AbstractSqlStore) ListDirectoryPrefixedEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, prefix string, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+
+	db, bucket, shortPath, err := store.getTxOrDB(ctx, dirPath, true)
+	if err != nil {
+		return lastFileName, fmt.Errorf("findDB %s : %v", dirPath, err)
 	}

-	rows, err := store.getTxOrDB(ctx).QueryContext(ctx, sqlText, util.HashStringToLong(string(fullpath)), startFileName, string(fullpath), prefix+"%", limit)
+	sqlText := store.GetSqlListExclusive(bucket)
+	if includeStartFile {
+		sqlText = store.GetSqlListInclusive(bucket)
+	}
+
+	rows, err := db.QueryContext(ctx, sqlText, util.HashStringToLong(string(shortPath)), startFileName, string(shortPath), prefix+"%", limit+1)
 	if err != nil {
-		return nil, fmt.Errorf("list %s : %v", fullpath, err)
+		return lastFileName, fmt.Errorf("list %s : %v", dirPath, err)
 	}
 	defer rows.Close()

@@ -187,28 +288,52 @@ func (store *AbstractSqlStore) ListDirectoryPrefixedEntries(ctx context.Context,
 		var name string
 		var data []byte
 		if err = rows.Scan(&name, &data); err != nil {
-			glog.V(0).Infof("scan %s : %v", fullpath, err)
-			return nil, fmt.Errorf("scan %s: %v", fullpath, err)
+			glog.V(0).Infof("scan %s : %v", dirPath, err)
+			return lastFileName, fmt.Errorf("scan %s: %v", dirPath, err)
 		}
+		lastFileName = name

 		entry := &filer.Entry{
-			FullPath: util.NewFullPath(string(fullpath), name),
+			FullPath: util.NewFullPath(string(dirPath), name),
 		}
 		if err = entry.DecodeAttributesAndChunks(util.MaybeDecompressData(data)); err != nil {
 			glog.V(0).Infof("scan decode %s : %v", entry.FullPath, err)
-			return nil, fmt.Errorf("scan decode %s : %v", entry.FullPath, err)
+			return lastFileName, fmt.Errorf("scan decode %s : %v", entry.FullPath, err)
+		}
+
+		if !eachEntryFunc(entry) {
+			break
 		}

-		entries = append(entries, entry)
 	}

-	return entries, nil
+	return lastFileName, nil
 }

-func (store *AbstractSqlStore) ListDirectoryEntries(ctx context.Context, fullpath util.FullPath, startFileName string, inclusive bool, limit int) (entries []*filer.Entry, err error) {
-	return store.ListDirectoryPrefixedEntries(ctx, fullpath, startFileName, inclusive, limit, "")
+func (store *AbstractSqlStore) ListDirectoryEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+	return store.ListDirectoryPrefixedEntries(ctx, dirPath, startFileName, includeStartFile, limit, "", nil)
 }

 func (store *AbstractSqlStore) Shutdown() {
 	store.DB.Close()
 }
+
+func isValidBucket(bucket string) bool {
+	return bucket != DEFAULT_TABLE && bucket != ""
+}
+
+func (store *AbstractSqlStore) CreateTable(ctx context.Context, bucket string) error {
+	if !store.SupportBucketTable {
+		return nil
+	}
+	_, err := store.DB.ExecContext(ctx, store.SqlGenerator.GetSqlCreateTable(bucket))
+	return err
+}
+
+func (store *AbstractSqlStore) deleteTable(ctx context.Context, bucket string) error {
+	if !store.SupportBucketTable {
+		return nil
+	}
+	_, err := store.DB.ExecContext(ctx, store.SqlGenerator.GetSqlDropTable(bucket))
+	return err
+}
@@ -5,28 +5,35 @@ import (
 	"database/sql"
 	"encoding/base64"
 	"fmt"
-	"strings"
-
 	"github.com/chrislusf/seaweedfs/weed/filer"
 	"github.com/chrislusf/seaweedfs/weed/glog"
 	"github.com/chrislusf/seaweedfs/weed/util"
+	"strings"
 )

 func (store *AbstractSqlStore) KvPut(ctx context.Context, key []byte, value []byte) (err error) {

+	db, _, _, err := store.getTxOrDB(ctx, "", false)
+	if err != nil {
+		return fmt.Errorf("findDB: %v", err)
+	}
+
 	dirStr, dirHash, name := genDirAndName(key)

-	res, err := store.getTxOrDB(ctx).ExecContext(ctx, store.SqlInsert, dirHash, name, dirStr, value)
-	if err != nil {
-		if !strings.Contains(strings.ToLower(err.Error()), "duplicate") {
-			return fmt.Errorf("kv insert: %s", err)
-		}
+	res, err := db.ExecContext(ctx, store.GetSqlInsert(DEFAULT_TABLE), dirHash, name, dirStr, value)
+	if err == nil {
+		return
+	}
+
+	if !strings.Contains(strings.ToLower(err.Error()), "duplicate") {
+		// return fmt.Errorf("kv insert: %s", err)
+		// skip this since the error can be in a different language
 	}

 	// now the insert failed possibly due to duplication constraints
 	glog.V(1).Infof("kv insert falls back to update: %s", err)

-	res, err = store.getTxOrDB(ctx).ExecContext(ctx, store.SqlUpdate, value, dirHash, name, dirStr)
+	res, err = db.ExecContext(ctx, store.GetSqlUpdate(DEFAULT_TABLE), value, dirHash, name, dirStr)
 	if err != nil {
 		return fmt.Errorf("kv upsert: %s", err)
 	}
@@ -41,8 +48,13 @@ func (store *AbstractSqlStore) KvPut(ctx context.Context, key []byte, value []by

 func (store *AbstractSqlStore) KvGet(ctx context.Context, key []byte) (value []byte, err error) {

+	db, _, _, err := store.getTxOrDB(ctx, "", false)
+	if err != nil {
+		return nil, fmt.Errorf("findDB: %v", err)
+	}
+
 	dirStr, dirHash, name := genDirAndName(key)
-	row := store.getTxOrDB(ctx).QueryRowContext(ctx, store.SqlFind, dirHash, name, dirStr)
+	row := db.QueryRowContext(ctx, store.GetSqlFind(DEFAULT_TABLE), dirHash, name, dirStr)

 	err = row.Scan(&value)

@@ -59,9 +71,14 @@ func (store *AbstractSqlStore) KvGet(ctx context.Context, key []byte) (value []b

 func (store *AbstractSqlStore) KvDelete(ctx context.Context, key []byte) (err error) {

+	db, _, _, err := store.getTxOrDB(ctx, "", false)
+	if err != nil {
+		return fmt.Errorf("findDB: %v", err)
+	}
+
 	dirStr, dirHash, name := genDirAndName(key)

-	res, err := store.getTxOrDB(ctx).ExecContext(ctx, store.SqlDelete, dirHash, name, dirStr)
+	res, err := db.ExecContext(ctx, store.GetSqlDelete(DEFAULT_TABLE), dirHash, name, dirStr)
 	if err != nil {
 		return fmt.Errorf("kv delete: %s", err)
 	}
@@ -16,8 +16,9 @@ func init() {
 }

 type CassandraStore struct {
-	cluster *gocql.ClusterConfig
-	session *gocql.Session
+	cluster                 *gocql.ClusterConfig
+	session                 *gocql.Session
+	superLargeDirectoryHash map[string]string
 }

 func (store *CassandraStore) GetName() string {
@@ -30,10 +31,16 @@ func (store *CassandraStore) Initialize(configuration util.Configuration, prefix
 		configuration.GetStringSlice(prefix+"hosts"),
 		configuration.GetString(prefix+"username"),
 		configuration.GetString(prefix+"password"),
+		configuration.GetStringSlice(prefix+"superLargeDirectories"),
 	)
 }

-func (store *CassandraStore) initialize(keyspace string, hosts []string, username string, password string) (err error) {
+func (store *CassandraStore) isSuperLargeDirectory(dir string) (dirHash string, isSuperLargeDirectory bool) {
+	dirHash, isSuperLargeDirectory = store.superLargeDirectoryHash[dir]
+	return
+}
+
+func (store *CassandraStore) initialize(keyspace string, hosts []string, username string, password string, superLargeDirectories []string) (err error) {
 	store.cluster = gocql.NewCluster(hosts...)
 	if username != "" && password != "" {
 		store.cluster.Authenticator = gocql.PasswordAuthenticator{Username: username, Password: password}
@@ -44,6 +51,19 @@ func (store *CassandraStore) initialize(keyspace string, hosts []string, usernam
 	if err != nil {
 		glog.V(0).Infof("Failed to open cassandra store, hosts %v, keyspace %s", hosts, keyspace)
 	}
+
+	// set directory hash
+	store.superLargeDirectoryHash = make(map[string]string)
+	existingHash := make(map[string]string)
+	for _, dir := range superLargeDirectories {
+		// adding dir hash to avoid duplicated names
+		dirHash := util.Md5String([]byte(dir))[:4]
+		store.superLargeDirectoryHash[dir] = dirHash
+		if existingDir, found := existingHash[dirHash]; found {
+			glog.Fatalf("directory %s has the same hash as %s", dir, existingDir)
+		}
+		existingHash[dirHash] = dir
+	}
 	return
 }

@@ -60,6 +80,10 @@ func (store *CassandraStore) RollbackTransaction(ctx context.Context) error {
 func (store *CassandraStore) InsertEntry(ctx context.Context, entry *filer.Entry) (err error) {

 	dir, name := entry.FullPath.DirAndName()
+	if dirHash, ok := store.isSuperLargeDirectory(dir); ok {
+		dir, name = dirHash+name, ""
+	}
+
 	meta, err := entry.EncodeAttributesAndChunks()
 	if err != nil {
 		return fmt.Errorf("encode %s: %s", entry.FullPath, err)
@@ -86,6 +110,10 @@ func (store *CassandraStore) UpdateEntry(ctx context.Context, entry *filer.Entry
 func (store *CassandraStore) FindEntry(ctx context.Context, fullpath util.FullPath) (entry *filer.Entry, err error) {

 	dir, name := fullpath.DirAndName()
+	if dirHash, ok := store.isSuperLargeDirectory(dir); ok {
+		dir, name = dirHash+name, ""
+	}
+
 	var data []byte
 	if err := store.session.Query(
 		"SELECT meta FROM filemeta WHERE directory=? AND name=?",
@@ -113,6 +141,9 @@ func (store *CassandraStore) FindEntry(ctx context.Context, fullpath util.FullPa
 func (store *CassandraStore) DeleteEntry(ctx context.Context, fullpath util.FullPath) error {

 	dir, name := fullpath.DirAndName()
+	if dirHash, ok := store.isSuperLargeDirectory(dir); ok {
+		dir, name = dirHash+name, ""
+	}

 	if err := store.session.Query(
 		"DELETE FROM filemeta WHERE directory=? AND name=?",
@@ -124,6 +155,9 @@ func (store *CassandraStore) DeleteEntry(ctx context.Context, fullpath util.Full
 }

 func (store *CassandraStore) DeleteFolderChildren(ctx context.Context, fullpath util.FullPath) error {
+	if _, ok := store.isSuperLargeDirectory(string(fullpath)); ok {
+		return nil // filer.ErrUnsupportedSuperLargeDirectoryListing
+	}

 	if err := store.session.Query(
 		"DELETE FROM filemeta WHERE directory=?",
@@ -134,37 +168,43 @@ func (store *CassandraStore) DeleteFolderChildren(ctx context.Context, fullpath
 	return nil
 }

-func (store *CassandraStore) ListDirectoryPrefixedEntries(ctx context.Context, fullpath util.FullPath, startFileName string, inclusive bool, limit int, prefix string) (entries []*filer.Entry, err error) {
-	return nil, filer.ErrUnsupportedListDirectoryPrefixed
+func (store *CassandraStore) ListDirectoryPrefixedEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, prefix string, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+	return lastFileName, filer.ErrUnsupportedListDirectoryPrefixed
 }

-func (store *CassandraStore) ListDirectoryEntries(ctx context.Context, fullpath util.FullPath, startFileName string, inclusive bool,
-	limit int) (entries []*filer.Entry, err error) {
+func (store *CassandraStore) ListDirectoryEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+
+	if _, ok := store.isSuperLargeDirectory(string(dirPath)); ok {
+		return // nil, filer.ErrUnsupportedSuperLargeDirectoryListing
+	}

 	cqlStr := "SELECT NAME, meta FROM filemeta WHERE directory=? AND name>? ORDER BY NAME ASC LIMIT ?"
-	if inclusive {
+	if includeStartFile {
 		cqlStr = "SELECT NAME, meta FROM filemeta WHERE directory=? AND name>=? ORDER BY NAME ASC LIMIT ?"
 	}

 	var data []byte
 	var name string
-	iter := store.session.Query(cqlStr, string(fullpath), startFileName, limit).Iter()
+	iter := store.session.Query(cqlStr, string(dirPath), startFileName, limit+1).Iter()
 	for iter.Scan(&name, &data) {
 		entry := &filer.Entry{
-			FullPath: util.NewFullPath(string(fullpath), name),
+			FullPath: util.NewFullPath(string(dirPath), name),
 		}
+		lastFileName = name
 		if decodeErr := entry.DecodeAttributesAndChunks(util.MaybeDecompressData(data)); decodeErr != nil {
 			err = decodeErr
 			glog.V(0).Infof("list %s : %v", entry.FullPath, err)
 			break
 		}
-		entries = append(entries, entry)
+		if !eachEntryFunc(entry) {
+			break
+		}
 	}
 	if err := iter.Close(); err != nil {
 		glog.V(0).Infof("list iterator close: %v", err)
 	}

-	return entries, err
+	return lastFileName, err
 }

 func (store *CassandraStore) Shutdown() {
@@ -1,42 +1,85 @@
 package filer

 import (
-	"os"
-
 	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/spf13/viper"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"os"
+	"reflect"
+	"strings"
 )

 var (
 	Stores []FilerStore
 )

-func (f *Filer) LoadConfiguration(config *viper.Viper) {
+func (f *Filer) LoadConfiguration(config *util.ViperProxy) {

 	validateOneEnabledStore(config)

+	// load configuration for default filer store
+	hasDefaultStoreConfigured := false
 	for _, store := range Stores {
 		if config.GetBool(store.GetName() + ".enabled") {
+			store = reflect.New(reflect.ValueOf(store).Elem().Type()).Interface().(FilerStore)
 			if err := store.Initialize(config, store.GetName()+"."); err != nil {
-				glog.Fatalf("Failed to initialize store for %s: %+v",
-					store.GetName(), err)
+				glog.Fatalf("failed to initialize store for %s: %+v", store.GetName(), err)
 			}
 			f.SetStore(store)
-			glog.V(0).Infof("Configure filer for %s", store.GetName())
-			return
+			glog.V(0).Infof("configured filer store to %s", store.GetName())
+			hasDefaultStoreConfigured = true
+			break
 		}
 	}

-	println()
-	println("Supported filer stores are:")
+	if !hasDefaultStoreConfigured {
+		println()
+		println("Supported filer stores are:")
+		for _, store := range Stores {
+			println("    " + store.GetName())
+		}
+		os.Exit(-1)
+	}
+
+	// load path-specific filer store here
+	// f.Store.AddPathSpecificStore(path, store)
+	storeNames := make(map[string]FilerStore)
 	for _, store := range Stores {
-		println("    " + store.GetName())
+		storeNames[store.GetName()] = store
+	}
+	allKeys := config.AllKeys()
+	for _, key := range allKeys {
+		if !strings.HasSuffix(key, ".enabled") {
+			continue
+		}
+		key = key[:len(key)-len(".enabled")]
+		if !strings.Contains(key, ".") {
+			continue
+		}
+
+		parts := strings.Split(key, ".")
+		storeName, storeId := parts[0], parts[1]
+
+		store, found := storeNames[storeName]
+		if !found {
+			continue
+		}
+		store = reflect.New(reflect.ValueOf(store).Elem().Type()).Interface().(FilerStore)
+		if err := store.Initialize(config, key+"."); err != nil {
+			glog.Fatalf("Failed to initialize store for %s: %+v", key, err)
+		}
+		location := config.GetString(key + ".location")
+		if location == "" {
+			glog.Errorf("path-specific filer store needs %s", key+".location")
+			os.Exit(-1)
+		}
+		f.Store.AddPathSpecificStore(location, storeId, store)
+
+		glog.V(0).Infof("configure filer %s for %s", store.GetName(), location)
 	}

-	os.Exit(-1)
 }

-func validateOneEnabledStore(config *viper.Viper) {
+func validateOneEnabledStore(config *util.ViperProxy) {
 	enabledStore := ""
 	for _, store := range Stores {
 		if config.GetBool(store.GetName() + ".enabled") {
@@ -96,8 +96,8 @@ func (store *ElasticStore) RollbackTransaction(ctx context.Context) error {
 	return nil
 }

-func (store *ElasticStore) ListDirectoryPrefixedEntries(ctx context.Context, fullpath weed_util.FullPath, startFileName string, inclusive bool, limit int, prefix string) (entries []*filer.Entry, err error) {
-	return nil, filer.ErrUnsupportedListDirectoryPrefixed
+func (store *ElasticStore) ListDirectoryPrefixedEntries(ctx context.Context, dirPath weed_util.FullPath, startFileName string, includeStartFile bool, limit int64, prefix string, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+	return lastFileName, filer.ErrUnsupportedListDirectoryPrefixed
 }

 func (store *ElasticStore) InsertEntry(ctx context.Context, entry *filer.Entry) (err error) {
@@ -187,28 +187,28 @@ func (store *ElasticStore) deleteEntry(ctx context.Context, index, id string) (e
 }

 func (store *ElasticStore) DeleteFolderChildren(ctx context.Context, fullpath weed_util.FullPath) (err error) {
-	if entries, err := store.ListDirectoryEntries(ctx, fullpath, "", false, math.MaxInt32); err == nil {
-		for _, entry := range entries {
-			store.DeleteEntry(ctx, entry.FullPath)
+	_, err = store.ListDirectoryEntries(ctx, fullpath, "", false, math.MaxInt32, func(entry *filer.Entry) bool {
+		if err := store.DeleteEntry(ctx, entry.FullPath); err != nil {
+			glog.Errorf("elastic delete %s: %v.", entry.FullPath, err)
+			return false
 		}
-	}
-	return nil
+		return true
+	})
+	return
 }

-func (store *ElasticStore) ListDirectoryEntries(
-	ctx context.Context, fullpath weed_util.FullPath, startFileName string, inclusive bool, limit int,
-) (entries []*filer.Entry, err error) {
-	if string(fullpath) == "/" {
-		return store.listRootDirectoryEntries(ctx, startFileName, inclusive, limit)
+func (store *ElasticStore) ListDirectoryEntries(ctx context.Context, dirPath weed_util.FullPath, startFileName string, includeStartFile bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+	if string(dirPath) == "/" {
+		return store.listRootDirectoryEntries(ctx, startFileName, includeStartFile, limit, eachEntryFunc)
 	}
-	return store.listDirectoryEntries(ctx, fullpath, startFileName, inclusive, limit)
+	return store.listDirectoryEntries(ctx, dirPath, startFileName, includeStartFile, limit, eachEntryFunc)
 }

-func (store *ElasticStore) listRootDirectoryEntries(ctx context.Context, startFileName string, inclusive bool, limit int) (entries []*filer.Entry, err error) {
+func (store *ElasticStore) listRootDirectoryEntries(ctx context.Context, startFileName string, inclusive bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
 	indexResult, err := store.client.CatIndices().Do(ctx)
 	if err != nil {
 		glog.Errorf("list indices %v.", err)
-		return entries, err
+		return
 	}
 	for _, index := range indexResult {
 		if index.Index == indexKV {
@@ -218,6 +218,7 @@ func (store *ElasticStore) listRootDirectoryEntries(ctx context.Context, startFi
 			if entry, err := store.FindEntry(ctx,
 				weed_util.FullPath("/"+strings.Replace(index.Index, indexPrefix, "", 1))); err == nil {
 				fileName := getFileName(entry.FullPath)
+				lastFileName = fileName
 				if fileName == startFileName && !inclusive {
 					continue
 				}
@@ -225,24 +226,25 @@ func (store *ElasticStore) listRootDirectoryEntries(ctx context.Context, startFi
 				if limit < 0 {
 					break
 				}
-				entries = append(entries, entry)
+				if !eachEntryFunc(entry) {
+					break
+				}
 			}
 		}
 	}
-	return entries, nil
+	return
 }

 func (store *ElasticStore) listDirectoryEntries(
-	ctx context.Context, fullpath weed_util.FullPath, startFileName string, inclusive bool, limit int,
-) (entries []*filer.Entry, err error) {
+	ctx context.Context, fullpath weed_util.FullPath, startFileName string, inclusive bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
 	first := true
 	index := getIndex(fullpath)
 	nextStart := ""
 	parentId := weed_util.Md5String([]byte(fullpath))
-	if _, err := store.client.Refresh(index).Do(ctx); err != nil {
+	if _, err = store.client.Refresh(index).Do(ctx); err != nil {
 		if elastic.IsNotFound(err) {
 			store.client.CreateIndex(index).Do(ctx)
-			return entries, nil
+			return
 		}
 	}
 	for {
@@ -250,7 +252,7 @@ func (store *ElasticStore) listDirectoryEntries(
 		if (startFileName == "" && first) || inclusive {
 			if result, err = store.search(ctx, index, parentId); err != nil {
 				glog.Errorf("search (%s,%s,%t,%d) %v.", string(fullpath), startFileName, inclusive, limit, err)
-				return entries, err
+				return
 			}
 		} else {
 			fullPath := string(fullpath) + "/" + startFileName
@@ -260,7 +262,7 @@ func (store *ElasticStore) listDirectoryEntries(
 			after := weed_util.Md5String([]byte(fullPath))
 			if result, err = store.searchAfter(ctx, index, parentId, after); err != nil {
 				glog.Errorf("searchAfter (%s,%s,%t,%d) %v.", string(fullpath), startFileName, inclusive, limit, err)
-				return entries, err
+				return
 			}
 		}
 		first = false
@@ -272,21 +274,21 @@ func (store *ElasticStore) listDirectoryEntries(
 			if err := jsoniter.Unmarshal(hit.Source, esEntry); err == nil {
 				limit--
 				if limit < 0 {
-					return entries, nil
+					return lastFileName, nil
 				}
 				nextStart = string(esEntry.Entry.FullPath)
 				fileName := getFileName(esEntry.Entry.FullPath)
+				lastFileName = fileName
 				if fileName == startFileName && !inclusive {
 					continue
 				}
-				entries = append(entries, esEntry.Entry)
+				if !eachEntryFunc(esEntry.Entry) {
+					break
+				}
 			}
 		}
-		if len(result.Hits.Hits) < store.maxPageSize {
-			break
-		}
 	}
-	return entries, nil
+	return
 }

 func (store *ElasticStore) search(ctx context.Context, index, parentId string) (result *elastic.SearchResult, err error) {
@@ -40,10 +40,11 @@ type Entry struct {

 	HardLinkId      HardLinkId
 	HardLinkCounter int32
+	Content         []byte
 }

 func (entry *Entry) Size() uint64 {
-	return maxUint64(TotalSize(entry.Chunks), entry.FileSize)
+	return maxUint64(maxUint64(TotalSize(entry.Chunks), entry.FileSize), uint64(len(entry.Content)))
 }

 func (entry *Entry) Timestamp() time.Time {
@@ -66,6 +67,7 @@ func (entry *Entry) ToProtoEntry() *filer_pb.Entry {
 		Extended:        entry.Extended,
 		HardLinkId:      entry.HardLinkId,
 		HardLinkCounter: entry.HardLinkCounter,
+		Content:         entry.Content,
 	}
 }

@@ -98,6 +100,7 @@ func FromPbEntry(dir string, entry *filer_pb.Entry) *Entry {
 		Chunks:          entry.Chunks,
 		HardLinkId:      HardLinkId(entry.HardLinkId),
 		HardLinkCounter: entry.HardLinkCounter,
+		Content:         entry.Content,
 	}
 }

@@ -18,6 +18,7 @@ func (entry *Entry) EncodeAttributesAndChunks() ([]byte, error) {
 		Extended:        entry.Extended,
 		HardLinkId:      entry.HardLinkId,
 		HardLinkCounter: entry.HardLinkCounter,
+		Content:         entry.Content,
 	}
 	return proto.Marshal(message)
 }
@@ -38,6 +39,7 @@ func (entry *Entry) DecodeAttributesAndChunks(blob []byte) error {

 	entry.HardLinkId = message.HardLinkId
 	entry.HardLinkCounter = message.HardLinkCounter
+	entry.Content = message.Content

 	return nil
 }
@@ -122,6 +124,9 @@ func EqualEntry(a, b *Entry) bool {
 	if a.HardLinkCounter != b.HardLinkCounter {
 		return false
 	}
+	if !bytes.Equal(a.Content, b.Content) {
+		return false
+	}
 	return true
 }

@@ -1,6 +1,7 @@
 package etcd

 import (
+	"bytes"
 	"context"
 	"fmt"
 	"strings"
@@ -101,7 +102,7 @@ func (store *EtcdStore) FindEntry(ctx context.Context, fullpath weed_util.FullPa

 	resp, err := store.client.Get(ctx, string(key))
 	if err != nil {
-		return nil, fmt.Errorf("get %s : %v", entry.FullPath, err)
+		return nil, fmt.Errorf("get %s : %v", fullpath, err)
 	}

 	if len(resp.Kvs) == 0 {
@@ -139,25 +140,32 @@ func (store *EtcdStore) DeleteFolderChildren(ctx context.Context, fullpath weed_
 	return nil
 }

-func (store *EtcdStore) ListDirectoryPrefixedEntries(ctx context.Context, fullpath weed_util.FullPath, startFileName string, inclusive bool, limit int, prefix string) (entries []*filer.Entry, err error) {
-	return nil, filer.ErrUnsupportedListDirectoryPrefixed
+func (store *EtcdStore) ListDirectoryPrefixedEntries(ctx context.Context, dirPath weed_util.FullPath, startFileName string, includeStartFile bool, limit int64, prefix string, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+	return lastFileName, filer.ErrUnsupportedListDirectoryPrefixed
 }

-func (store *EtcdStore) ListDirectoryEntries(ctx context.Context, fullpath weed_util.FullPath, startFileName string, inclusive bool, limit int) (entries []*filer.Entry, err error) {
-	directoryPrefix := genDirectoryKeyPrefix(fullpath, "")
+func (store *EtcdStore) ListDirectoryEntries(ctx context.Context, dirPath weed_util.FullPath, startFileName string, includeStartFile bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+	directoryPrefix := genDirectoryKeyPrefix(dirPath, "")
+	lastFileStart := directoryPrefix
+	if startFileName != "" {
+		lastFileStart = genDirectoryKeyPrefix(dirPath, startFileName)
+	}

-	resp, err := store.client.Get(ctx, string(directoryPrefix),
+	resp, err := store.client.Get(ctx, string(lastFileStart),
 		clientv3.WithPrefix(), clientv3.WithSort(clientv3.SortByKey, clientv3.SortDescend))
 	if err != nil {
-		return nil, fmt.Errorf("list %s : %v", fullpath, err)
+		return lastFileName, fmt.Errorf("list %s : %v", dirPath, err)
 	}

 	for _, kv := range resp.Kvs {
+		if !bytes.HasPrefix(kv.Key, directoryPrefix) {
+			break
+		}
 		fileName := getNameFromKey(kv.Key)
 		if fileName == "" {
 			continue
 		}
-		if fileName == startFileName && !inclusive {
+		if fileName == startFileName && !includeStartFile {
 			continue
 		}
 		limit--
@@ -165,17 +173,20 @@ func (store *EtcdStore) ListDirectoryEntries(ctx context.Context, fullpath weed_
 			break
 		}
 		entry := &filer.Entry{
-			FullPath: weed_util.NewFullPath(string(fullpath), fileName),
+			FullPath: weed_util.NewFullPath(string(dirPath), fileName),
 		}
 		if decodeErr := entry.DecodeAttributesAndChunks(weed_util.MaybeDecompressData(kv.Value)); decodeErr != nil {
 			err = decodeErr
 			glog.V(0).Infof("list %s : %v", entry.FullPath, err)
 			break
 		}
-		entries = append(entries, entry)
+		if !eachEntryFunc(entry) {
+			break
+		}
+		lastFileName = fileName
 	}

-	return entries, err
+	return lastFileName, err
 }

 func genKey(dirPath, fileName string) (key []byte) {
@@ -3,6 +3,7 @@ package filer
 import (
 	"bytes"
 	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/wdclient"
 	"io"
 	"math"
 	"time"
@@ -38,7 +39,7 @@ func SeparateManifestChunks(chunks []*filer_pb.FileChunk) (manifestChunks, nonMa
 	return
 }

-func ResolveChunkManifest(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (dataChunks, manifestChunks []*filer_pb.FileChunk, manifestResolveErr error) {
+func ResolveChunkManifest(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (dataChunks, manifestChunks []*filer_pb.FileChunk, manifestResolveErr error) {
 	// TODO maybe parallel this
 	for _, chunk := range chunks {
 		if !chunk.IsChunkManifest {
@@ -63,7 +64,7 @@ func ResolveChunkManifest(lookupFileIdFn LookupFileIdFunctionType, chunks []*fil
 	return
 }

-func ResolveOneChunkManifest(lookupFileIdFn LookupFileIdFunctionType, chunk *filer_pb.FileChunk) (dataChunks []*filer_pb.FileChunk, manifestResolveErr error) {
+func ResolveOneChunkManifest(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunk *filer_pb.FileChunk) (dataChunks []*filer_pb.FileChunk, manifestResolveErr error) {
 	if !chunk.IsChunkManifest {
 		return
 	}
@@ -84,7 +85,7 @@ func ResolveOneChunkManifest(lookupFileIdFn LookupFileIdFunctionType, chunk *fil
 }

 // TODO fetch from cache for weed mount?
-func fetchChunk(lookupFileIdFn LookupFileIdFunctionType, fileId string, cipherKey []byte, isGzipped bool) ([]byte, error) {
+func fetchChunk(lookupFileIdFn wdclient.LookupFileIdFunctionType, fileId string, cipherKey []byte, isGzipped bool) ([]byte, error) {
 	urlStrings, err := lookupFileIdFn(fileId)
 	if err != nil {
 		glog.Errorf("operation LookupFileId %s failed, err: %v", fileId, err)
@@ -4,6 +4,7 @@ import (
 	"bytes"
 	"encoding/hex"
 	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/wdclient"
 	"math"
 	"sort"
 	"sync"
@@ -52,7 +53,7 @@ func ETagChunks(chunks []*filer_pb.FileChunk) (etag string) {
 	return fmt.Sprintf("%x-%d", util.Md5(bytes.Join(md5_digests, nil)), len(chunks))
 }

-func CompactFileChunks(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (compacted, garbage []*filer_pb.FileChunk) {
+func CompactFileChunks(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (compacted, garbage []*filer_pb.FileChunk) {

 	visibles, _ := NonOverlappingVisibleIntervals(lookupFileIdFn, chunks)

@@ -71,7 +72,7 @@ func CompactFileChunks(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_
 	return
 }

-func MinusChunks(lookupFileIdFn LookupFileIdFunctionType, as, bs []*filer_pb.FileChunk) (delta []*filer_pb.FileChunk, err error) {
+func MinusChunks(lookupFileIdFn wdclient.LookupFileIdFunctionType, as, bs []*filer_pb.FileChunk) (delta []*filer_pb.FileChunk, err error) {

 	aData, aMeta, aErr := ResolveChunkManifest(lookupFileIdFn, as)
 	if aErr != nil {
@@ -116,7 +117,7 @@ func (cv *ChunkView) IsFullChunk() bool {
 	return cv.Size == cv.ChunkSize
 }

-func ViewFromChunks(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_pb.FileChunk, offset int64, size int64) (views []*ChunkView) {
+func ViewFromChunks(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk, offset int64, size int64) (views []*ChunkView) {

 	visibles, _ := NonOverlappingVisibleIntervals(lookupFileIdFn, chunks)

@@ -222,7 +223,7 @@ func MergeIntoVisibles(visibles []VisibleInterval, chunk *filer_pb.FileChunk) (n

 // NonOverlappingVisibleIntervals translates the file chunk into VisibleInterval in memory
 // If the file chunk content is a chunk manifest
-func NonOverlappingVisibleIntervals(lookupFileIdFn LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (visibles []VisibleInterval, err error) {
+func NonOverlappingVisibleIntervals(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) (visibles []VisibleInterval, err error) {

 	chunks, _, err = ResolveChunkManifest(lookupFileIdFn, chunks)

@@ -18,7 +18,7 @@ import (

 const (
 	LogFlushInterval = time.Minute
-	PaginationSize   = 1024 * 256
+	PaginationSize   = 1024
 	FilerStoreId     = "filer.store.id"
 )

@@ -134,69 +134,7 @@ func (f *Filer) CreateEntry(ctx context.Context, entry *Entry, o_excl bool, isFr
 		return nil
 	}

-	dirParts := strings.Split(string(entry.FullPath), "/")
-
-	// fmt.Printf("directory parts: %+v\n", dirParts)
-
-	var lastDirectoryEntry *Entry
-
-	for i := 1; i < len(dirParts); i++ {
-		dirPath := "/" + util.Join(dirParts[:i]...)
-		// fmt.Printf("%d directory: %+v\n", i, dirPath)
-
-		// check the store directly
-		glog.V(4).Infof("find uncached directory: %s", dirPath)
-		dirEntry, _ := f.FindEntry(ctx, util.FullPath(dirPath))
-
-		// no such existing directory
-		if dirEntry == nil {
-
-			// create the directory
-			now := time.Now()
-
-			dirEntry = &Entry{
-				FullPath: util.FullPath(dirPath),
-				Attr: Attr{
-					Mtime:       now,
-					Crtime:      now,
-					Mode:        os.ModeDir | entry.Mode | 0110,
-					Uid:         entry.Uid,
-					Gid:         entry.Gid,
-					Collection:  entry.Collection,
-					Replication: entry.Replication,
-					UserName:    entry.UserName,
-					GroupNames:  entry.GroupNames,
-				},
-			}
-
-			glog.V(2).Infof("create directory: %s %v", dirPath, dirEntry.Mode)
-			mkdirErr := f.Store.InsertEntry(ctx, dirEntry)
-			if mkdirErr != nil {
-				if _, err := f.FindEntry(ctx, util.FullPath(dirPath)); err == filer_pb.ErrNotFound {
-					glog.V(3).Infof("mkdir %s: %v", dirPath, mkdirErr)
-					return fmt.Errorf("mkdir %s: %v", dirPath, mkdirErr)
-				}
-			} else {
-				f.maybeAddBucket(dirEntry)
-				f.NotifyUpdateEvent(ctx, nil, dirEntry, false, isFromOtherCluster, nil)
-			}
-
-		} else if !dirEntry.IsDirectory() {
-			glog.Errorf("CreateEntry %s: %s should be a directory", entry.FullPath, dirPath)
-			return fmt.Errorf("%s is a file", dirPath)
-		}
-
-		// remember the direct parent directory entry
-		if i == len(dirParts)-1 {
-			lastDirectoryEntry = dirEntry
-		}
-
-	}
-
-	if lastDirectoryEntry == nil {
-		glog.Errorf("CreateEntry %s: lastDirectoryEntry is nil", entry.FullPath)
-		return fmt.Errorf("parent folder not found: %v", entry.FullPath)
-	}
+	oldEntry, _ := f.FindEntry(ctx, entry.FullPath)

 	/*
 		if !hasWritePermission(lastDirectoryEntry, entry) {
@@ -206,9 +144,13 @@ func (f *Filer) CreateEntry(ctx context.Context, entry *Entry, o_excl bool, isFr
 		}
 	*/

-	oldEntry, _ := f.FindEntry(ctx, entry.FullPath)
-
 	if oldEntry == nil {
+
+		dirParts := strings.Split(string(entry.FullPath), "/")
+		if err := f.ensureParentDirecotryEntry(ctx, entry, dirParts, len(dirParts)-1, isFromOtherCluster); err != nil {
+			return err
+		}
+
 		glog.V(4).Infof("InsertEntry %s: new entry: %v", entry.FullPath, entry.Name())
 		if err := f.Store.InsertEntry(ctx, entry); err != nil {
 			glog.Errorf("insert entry %s: %v", entry.FullPath, err)
@@ -236,8 +178,68 @@ func (f *Filer) CreateEntry(ctx context.Context, entry *Entry, o_excl bool, isFr
 	return nil
 }

+func (f *Filer) ensureParentDirecotryEntry(ctx context.Context, entry *Entry, dirParts []string, level int, isFromOtherCluster bool) (err error) {
+
+	if level == 0 {
+		return nil
+	}
+
+	dirPath := "/" + util.Join(dirParts[:level]...)
+	// fmt.Printf("%d directory: %+v\n", i, dirPath)
+
+	// check the store directly
+	glog.V(4).Infof("find uncached directory: %s", dirPath)
+	dirEntry, _ := f.FindEntry(ctx, util.FullPath(dirPath))
+
+	// no such existing directory
+	if dirEntry == nil {
+
+		// ensure parent directory
+		if err = f.ensureParentDirecotryEntry(ctx, entry, dirParts, level-1, isFromOtherCluster); err != nil {
+			return err
+		}
+
+		// create the directory
+		now := time.Now()
+
+		dirEntry = &Entry{
+			FullPath: util.FullPath(dirPath),
+			Attr: Attr{
+				Mtime:       now,
+				Crtime:      now,
+				Mode:        os.ModeDir | entry.Mode | 0110,
+				Uid:         entry.Uid,
+				Gid:         entry.Gid,
+				Collection:  entry.Collection,
+				Replication: entry.Replication,
+				UserName:    entry.UserName,
+				GroupNames:  entry.GroupNames,
+			},
+		}
+
+		glog.V(2).Infof("create directory: %s %v", dirPath, dirEntry.Mode)
+		mkdirErr := f.Store.InsertEntry(ctx, dirEntry)
+		if mkdirErr != nil {
+			if _, err := f.FindEntry(ctx, util.FullPath(dirPath)); err == filer_pb.ErrNotFound {
+				glog.V(3).Infof("mkdir %s: %v", dirPath, mkdirErr)
+				return fmt.Errorf("mkdir %s: %v", dirPath, mkdirErr)
+			}
+		} else {
+			f.maybeAddBucket(dirEntry)
+			f.NotifyUpdateEvent(ctx, nil, dirEntry, false, isFromOtherCluster, nil)
+		}
+
+	} else if !dirEntry.IsDirectory() {
+		glog.Errorf("CreateEntry %s: %s should be a directory", entry.FullPath, dirPath)
+		return fmt.Errorf("%s is a file", dirPath)
+	}
+
+	return nil
+}
+
 func (f *Filer) UpdateEntry(ctx context.Context, oldEntry, entry *Entry) (err error) {
 	if oldEntry != nil {
+		entry.Attr.Crtime = oldEntry.Attr.Crtime
 		if oldEntry.IsDirectory() && !entry.IsDirectory() {
 			glog.Errorf("existing %s is a directory", entry.FullPath)
 			return fmt.Errorf("existing %s is a directory", entry.FullPath)
@@ -250,26 +252,28 @@ func (f *Filer) UpdateEntry(ctx context.Context, oldEntry, entry *Entry) (err er
 	return f.Store.UpdateEntry(ctx, entry)
 }

+var (
+	Root = &Entry{
+		FullPath: "/",
+		Attr: Attr{
+			Mtime:  time.Now(),
+			Crtime: time.Now(),
+			Mode:   os.ModeDir | 0755,
+			Uid:    OS_UID,
+			Gid:    OS_GID,
+		},
+	}
+)
+
 func (f *Filer) FindEntry(ctx context.Context, p util.FullPath) (entry *Entry, err error) {

-	now := time.Now()
-
 	if string(p) == "/" {
-		return &Entry{
-			FullPath: p,
-			Attr: Attr{
-				Mtime:  now,
-				Crtime: now,
-				Mode:   os.ModeDir | 0755,
-				Uid:    OS_UID,
-				Gid:    OS_GID,
-			},
-		}, nil
+		return Root, nil
 	}
 	entry, err = f.Store.FindEntry(ctx, p)
 	if entry != nil && entry.TtlSec > 0 {
 		if entry.Crtime.Add(time.Duration(entry.TtlSec) * time.Second).Before(time.Now()) {
-			f.Store.DeleteEntry(ctx, p.Child(entry.Name()))
+			f.Store.DeleteOneEntry(ctx, entry)
 			return nil, filer_pb.ErrNotFound
 		}
 	}
@@ -277,38 +281,19 @@ func (f *Filer) FindEntry(ctx context.Context, p util.FullPath) (entry *Entry, e

 }

-func (f *Filer) ListDirectoryEntries(ctx context.Context, p util.FullPath, startFileName string, inclusive bool, limit int, prefix string) ([]*Entry, error) {
-	if strings.HasSuffix(string(p), "/") && len(p) > 1 {
-		p = p[0 : len(p)-1]
-	}
-
-	var makeupEntries []*Entry
-	entries, expiredCount, lastFileName, err := f.doListDirectoryEntries(ctx, p, startFileName, inclusive, limit, prefix)
-	for expiredCount > 0 && err == nil {
-		makeupEntries, expiredCount, lastFileName, err = f.doListDirectoryEntries(ctx, p, lastFileName, false, expiredCount, prefix)
-		if err == nil {
-			entries = append(entries, makeupEntries...)
-		}
-	}
-
-	return entries, err
-}
-
-func (f *Filer) doListDirectoryEntries(ctx context.Context, p util.FullPath, startFileName string, inclusive bool, limit int, prefix string) (entries []*Entry, expiredCount int, lastFileName string, err error) {
-	listedEntries, listErr := f.Store.ListDirectoryPrefixedEntries(ctx, p, startFileName, inclusive, limit, prefix)
-	if listErr != nil {
-		return listedEntries, expiredCount, "", listErr
-	}
-	for _, entry := range listedEntries {
-		lastFileName = entry.Name()
+func (f *Filer) doListDirectoryEntries(ctx context.Context, p util.FullPath, startFileName string, inclusive bool, limit int64, prefix string, eachEntryFunc ListEachEntryFunc) (expiredCount int64, lastFileName string, err error) {
+	lastFileName, err = f.Store.ListDirectoryPrefixedEntries(ctx, p, startFileName, inclusive, limit, prefix, func(entry *Entry) bool {
 		if entry.TtlSec > 0 {
 			if entry.Crtime.Add(time.Duration(entry.TtlSec) * time.Second).Before(time.Now()) {
-				f.Store.DeleteEntry(ctx, p.Child(entry.Name()))
+				f.Store.DeleteOneEntry(ctx, entry)
 				expiredCount++
-				continue
+				return true
 			}
 		}
-		entries = append(entries, entry)
+		return eachEntryFunc(entry)
+	})
+	if err != nil {
+		return expiredCount, lastFileName, err
 	}
 	return
 }
@@ -317,11 +302,3 @@ func (f *Filer) Shutdown() {
 	f.LocalMetaLogBuffer.Shutdown()
 	f.Store.Shutdown()
 }
-
-func (f *Filer) maybeDeleteHardLinks(hardLinkIds []HardLinkId) {
-	for _, hardLinkId := range hardLinkIds {
-		if err := f.Store.DeleteHardLink(context.Background(), hardLinkId); err != nil {
-			glog.Errorf("delete hard link id %d : %v", hardLinkId, err)
-		}
-	}
-}
@@ -27,9 +27,9 @@ func (f *Filer) LoadBuckets() {
 		buckets: make(map[BucketName]*BucketOption),
 	}

-	limit := math.MaxInt32
+	limit := int64(math.MaxInt32)

-	entries, err := f.ListDirectoryEntries(context.Background(), util.FullPath(f.DirBucketsPath), "", false, limit, "")
+	entries, _, err := f.ListDirectoryEntries(context.Background(), util.FullPath(f.DirBucketsPath), "", false, limit, "", "")

 	if err != nil {
 		glog.V(1).Infof("no buckets found: %v", err)
@@ -9,13 +9,15 @@ import (
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
 	"github.com/chrislusf/seaweedfs/weed/util"
 	"github.com/golang/protobuf/jsonpb"
-	"github.com/golang/protobuf/proto"
 	"github.com/viant/ptrie"
 )

 const (
-	DirectoryEtc  = "/etc"
-	FilerConfName = "filer.conf"
+	DirectoryEtcRoot      = "/etc"
+	DirectoryEtcSeaweedFS = "/etc/seaweedfs"
+	FilerConfName         = "filer.conf"
+	IamConfigDirecotry    = "/etc/iam"
+	IamIdentityFile       = "identity.json"
 )

 type FilerConf struct {
@@ -30,7 +32,7 @@ func NewFilerConf() (fc *FilerConf) {
 }

 func (fc *FilerConf) loadFromFiler(filer *Filer) (err error) {
-	filerConfPath := util.NewFullPath(DirectoryEtc, FilerConfName)
+	filerConfPath := util.NewFullPath(DirectoryEtcSeaweedFS, FilerConfName)
 	entry, err := filer.FindEntry(context.Background(), filerConfPath)
 	if err != nil {
 		if err == filer_pb.ErrNotFound {
@@ -40,6 +42,10 @@ func (fc *FilerConf) loadFromFiler(filer *Filer) (err error) {
 		return
 	}

+	if len(entry.Content) > 0 {
+		return fc.LoadFromBytes(entry.Content)
+	}
+
 	return fc.loadFromChunks(filer, entry.Chunks)
 }

@@ -57,15 +63,7 @@ func (fc *FilerConf) LoadFromBytes(data []byte) (err error) {
 	conf := &filer_pb.FilerConf{}

 	if err := jsonpb.Unmarshal(bytes.NewReader(data), conf); err != nil {
-
-		err = proto.UnmarshalText(string(data), conf)
-		if err != nil {
-			glog.Errorf("unable to parse filer conf: %v", err)
-			// this is not recoverable
-			return nil
-		}
-
-		return nil
+		return err
 	}

 	return fc.doLoadConf(conf)
@@ -22,7 +22,7 @@ func TestFilerConf(t *testing.T) {
 		},
 		{
 			LocationPrefix: "/buckets/",
-			Replication: "001",
+			Replication:    "001",
 		},
 	}}
 	fc.doLoadConf(conf)
@@ -3,7 +3,6 @@ package filer
 import (
 	"context"
 	"fmt"
-
 	"github.com/chrislusf/seaweedfs/weed/glog"
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
 	"github.com/chrislusf/seaweedfs/weed/pb/master_pb"
@@ -22,7 +21,7 @@ func (f *Filer) DeleteEntryMetaAndData(ctx context.Context, p util.FullPath, isR
 		return findErr
 	}

-	isCollection := f.isBucket(entry)
+	isDeleteCollection := f.isBucket(entry)

 	var chunks []*filer_pb.FileChunk
 	var hardLinkIds []HardLinkId
@@ -31,7 +30,7 @@ func (f *Filer) DeleteEntryMetaAndData(ctx context.Context, p util.FullPath, isR
 		// delete the folder children, not including the folder itself
 		var dirChunks []*filer_pb.FileChunk
 		var dirHardLinkIds []HardLinkId
-		dirChunks, dirHardLinkIds, err = f.doBatchDeleteFolderMetaAndData(ctx, entry, isRecursive, ignoreRecursiveError, shouldDeleteChunks && !isCollection, isFromOtherCluster, signatures)
+		dirChunks, dirHardLinkIds, err = f.doBatchDeleteFolderMetaAndData(ctx, entry, isRecursive, ignoreRecursiveError, shouldDeleteChunks && !isDeleteCollection, isDeleteCollection, isFromOtherCluster, signatures)
 		if err != nil {
 			glog.V(0).Infof("delete directory %s: %v", p, err)
 			return fmt.Errorf("delete directory %s: %v", p, err)
@@ -46,7 +45,7 @@ func (f *Filer) DeleteEntryMetaAndData(ctx context.Context, p util.FullPath, isR
 		return fmt.Errorf("delete file %s: %v", p, err)
 	}

-	if shouldDeleteChunks && !isCollection {
+	if shouldDeleteChunks && !isDeleteCollection {
 		f.DirectDeleteChunks(chunks)
 	}
 	// A case not handled:
@@ -55,7 +54,7 @@ func (f *Filer) DeleteEntryMetaAndData(ctx context.Context, p util.FullPath, isR
 		f.maybeDeleteHardLinks(hardLinkIds)
 	}

-	if isCollection {
+	if isDeleteCollection {
 		collectionName := entry.Name()
 		f.doDeleteCollection(collectionName)
 		f.deleteBucket(collectionName)
@@ -64,46 +63,49 @@ func (f *Filer) DeleteEntryMetaAndData(ctx context.Context, p util.FullPath, isR
 	return nil
 }

-func (f *Filer) doBatchDeleteFolderMetaAndData(ctx context.Context, entry *Entry, isRecursive, ignoreRecursiveError, shouldDeleteChunks, isFromOtherCluster bool, signatures []int32) (chunks []*filer_pb.FileChunk, hardlinkIds []HardLinkId, err error) {
+func (f *Filer) doBatchDeleteFolderMetaAndData(ctx context.Context, entry *Entry, isRecursive, ignoreRecursiveError, shouldDeleteChunks, isDeletingBucket, isFromOtherCluster bool, signatures []int32) (chunks []*filer_pb.FileChunk, hardlinkIds []HardLinkId, err error) {

 	lastFileName := ""
 	includeLastFile := false
-	for {
-		entries, err := f.ListDirectoryEntries(ctx, entry.FullPath, lastFileName, includeLastFile, PaginationSize, "")
-		if err != nil {
-			glog.Errorf("list folder %s: %v", entry.FullPath, err)
-			return nil, nil, fmt.Errorf("list folder %s: %v", entry.FullPath, err)
-		}
-		if lastFileName == "" && !isRecursive && len(entries) > 0 {
-			// only for first iteration in the loop
-			glog.Errorf("deleting a folder %s has children: %+v ...", entry.FullPath, entries[0].Name())
-			return nil, nil, fmt.Errorf("fail to delete non-empty folder: %s", entry.FullPath)
-		}
+	if !isDeletingBucket {
+		for {
+			entries, _, err := f.ListDirectoryEntries(ctx, entry.FullPath, lastFileName, includeLastFile, PaginationSize, "", "")
+			if err != nil {
+				glog.Errorf("list folder %s: %v", entry.FullPath, err)
+				return nil, nil, fmt.Errorf("list folder %s: %v", entry.FullPath, err)
+			}
+			if lastFileName == "" && !isRecursive && len(entries) > 0 {
+				// only for first iteration in the loop
+				glog.Errorf("deleting a folder %s has children: %+v ...", entry.FullPath, entries[0].Name())
+				return nil, nil, fmt.Errorf("fail to delete non-empty folder: %s", entry.FullPath)
+			}

-		for _, sub := range entries {
-			lastFileName = sub.Name()
-			var dirChunks []*filer_pb.FileChunk
-			var dirHardLinkIds []HardLinkId
-			if sub.IsDirectory() {
-				dirChunks, dirHardLinkIds, err = f.doBatchDeleteFolderMetaAndData(ctx, sub, isRecursive, ignoreRecursiveError, shouldDeleteChunks, false, nil)
-				chunks = append(chunks, dirChunks...)
-				hardlinkIds = append(hardlinkIds, dirHardLinkIds...)
-			} else {
-				f.NotifyUpdateEvent(ctx, sub, nil, shouldDeleteChunks, isFromOtherCluster, nil)
-				if len(sub.HardLinkId) != 0 {
-					// hard link chunk data are deleted separately
-					hardlinkIds = append(hardlinkIds, sub.HardLinkId)
+			for _, sub := range entries {
+				lastFileName = sub.Name()
+				var dirChunks []*filer_pb.FileChunk
+				var dirHardLinkIds []HardLinkId
+				if sub.IsDirectory() {
+					subIsDeletingBucket := f.isBucket(sub)
+					dirChunks, dirHardLinkIds, err = f.doBatchDeleteFolderMetaAndData(ctx, sub, isRecursive, ignoreRecursiveError, shouldDeleteChunks, subIsDeletingBucket, false, nil)
+					chunks = append(chunks, dirChunks...)
+					hardlinkIds = append(hardlinkIds, dirHardLinkIds...)
 				} else {
-					chunks = append(chunks, sub.Chunks...)
+					f.NotifyUpdateEvent(ctx, sub, nil, shouldDeleteChunks, isFromOtherCluster, nil)
+					if len(sub.HardLinkId) != 0 {
+						// hard link chunk data are deleted separately
+						hardlinkIds = append(hardlinkIds, sub.HardLinkId)
+					} else {
+						chunks = append(chunks, sub.Chunks...)
+					}
+				}
+				if err != nil && !ignoreRecursiveError {
+					return nil, nil, err
 				}
 			}
-			if err != nil && !ignoreRecursiveError {
-				return nil, nil, err
-			}
-		}

-		if len(entries) < PaginationSize {
-			break
+			if len(entries) < PaginationSize {
+				break
+			}
 		}
 	}

@@ -122,7 +124,7 @@ func (f *Filer) doDeleteEntryMetaAndData(ctx context.Context, entry *Entry, shou

 	glog.V(3).Infof("deleting entry %v, delete chunks: %v", entry.FullPath, shouldDeleteChunks)

-	if storeDeletionErr := f.Store.DeleteEntry(ctx, entry.FullPath); storeDeletionErr != nil {
+	if storeDeletionErr := f.Store.DeleteOneEntry(ctx, entry); storeDeletionErr != nil {
 		return fmt.Errorf("filer store delete: %v", storeDeletionErr)
 	}
 	if !entry.IsDirectory() {
@@ -145,3 +147,11 @@ func (f *Filer) doDeleteCollection(collectionName string) (err error) {
 	})

 }
+
+func (f *Filer) maybeDeleteHardLinks(hardLinkIds []HardLinkId) {
+	for _, hardLinkId := range hardLinkIds {
+		if err := f.Store.DeleteHardLink(context.Background(), hardLinkId); err != nil {
+			glog.Errorf("delete hard link id %d : %v", hardLinkId, err)
+		}
+	}
+}
@@ -113,13 +113,13 @@ func (f *Filer) ReadPersistedLogBuffer(startTime time.Time, eachLogEntryFn func(
 	sizeBuf := make([]byte, 4)
 	startTsNs := startTime.UnixNano()

-	dayEntries, listDayErr := f.ListDirectoryEntries(context.Background(), SystemLogDir, startDate, true, 366, "")
+	dayEntries, _, listDayErr := f.ListDirectoryEntries(context.Background(), SystemLogDir, startDate, true, 366, "", "")
 	if listDayErr != nil {
 		return lastTsNs, fmt.Errorf("fail to list log by day: %v", listDayErr)
 	}
 	for _, dayEntry := range dayEntries {
 		// println("checking day", dayEntry.FullPath)
-		hourMinuteEntries, listHourMinuteErr := f.ListDirectoryEntries(context.Background(), util.NewFullPath(SystemLogDir, dayEntry.Name()), "", false, 24*60, "")
+		hourMinuteEntries, _, listHourMinuteErr := f.ListDirectoryEntries(context.Background(), util.NewFullPath(SystemLogDir, dayEntry.Name()), "", false, 24*60, "", "")
 		if listHourMinuteErr != nil {
 			return lastTsNs, fmt.Errorf("fail to list log %s by day: %v", dayEntry.Name(), listHourMinuteErr)
 		}
@@ -170,7 +170,7 @@ func ReadEachLogEntry(r io.Reader, sizeBuf []byte, ns int64, eachLogEntryFn func
 			return lastTsNs, err
 		}
 		if logEntry.TsNs <= ns {
-			return lastTsNs, nil
+			continue
 		}
 		// println("each log: ", logEntry.TsNs)
 		if err := eachLogEntryFn(logEntry); err != nil {
@@ -11,8 +11,8 @@ import (

 // onMetadataChangeEvent is triggered after filer processed change events from local or remote filers
 func (f *Filer) onMetadataChangeEvent(event *filer_pb.SubscribeMetadataResponse) {
-	if DirectoryEtc != event.Directory {
-		if DirectoryEtc != event.EventNotification.NewParentPath {
+	if DirectoryEtcSeaweedFS != event.Directory {
+		if DirectoryEtcSeaweedFS != event.EventNotification.NewParentPath {
 			return
 		}
 	}
--- a/Show More
+++ b/Show More