2.27

use basic file object to parse path
fix https://github.com/chrislusf/seaweedfs/issues/1825
2026-10-02 04:32:42 +00:00 · 2021-02-21 19:29:27 -08:00 · 2021-02-21 19:27:38 -08:00 · 2021-02-20 22:35:18 -08:00 · 2021-02-20 21:19:21 -08:00 · 2021-02-20 20:06:06 -08:00
376 changed files with 17860 additions and 7290 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: ${{ github.token }}
+          tag: dev
+          fail-if-no-assets: false
+          assets: |
+            weed-*
@@ -0,0 +1,62 @@
+name: Release
+
+on:
+  push:
+    branches: [ master ]
+
+jobs:
+
+  build:
+    name: Build
+    runs-on: ubuntu-latest
+    strategy:
+      matrix:
+        goos: [linux, windows, darwin, freebsd ]
+        goarch: [amd64, arm]
+        exclude:
+          - goarch: arm
+            goos: darwin
+          - goarch: arm
+            goos: windows
+
+    steps:
+
+      - name: Check out code into the Go module directory
+        uses: actions/checkout@v2
+
+      - name: Wait for the deletion
+        uses: jakejarvis/wait-action@master
+        with:
+          time: '30s'
+
+      - name: Set BUILD_TIME env
+        run: echo BUILD_TIME=$(date -u +%Y-%m-%d-%H-%M) >> ${GITHUB_ENV}
+
+      - name: Go Release Binaries
+        uses: wangyoucao577/go-release-action@v1.14
+        with:
+          github_token: ${{ secrets.GITHUB_TOKEN }}
+          goos: ${{ matrix.goos }}
+          goarch: ${{ matrix.goarch }}
+          release_tag: dev
+          overwrite: true
+          build_flags: -tags 5BytesOffset # optional, default is
+          ldflags: -extldflags -static -X github.com/chrislusf/seaweedfs/weed/util.COMMIT=${{github.sha}}
+          # 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@v1.14
+        with:
+          github_token: ${{ secrets.GITHUB_TOKEN }}
+          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-
+          asset_name: "weed-${{ env.BUILD_TIME }}-${{ matrix.goos }}-${{ matrix.goarch }}"
@@ -1,8 +1,8 @@
 sudo: false
 language: go
 go:
-  - 1.14.x
  - 1.15.x
+  - 1.16.x

 before_install:
  - export PATH=/home/travis/gopath/bin:$PATH
@@ -44,4 +44,4 @@ deploy:
  on:
    tags: true
    repo: chrislusf/seaweedfs
-    go: 1.15.x
+    go: 1.16.x
@@ -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)
+
 ---


@@ -58,11 +47,12 @@ Your support will be really appreciated by me and other supporters!
 Table of Contents
 =================

+* [Quick Start](#quick-start)
 * [Introduction](#introduction)
 * [Features](#features)
    * [Additional Features](#additional-features)
    * [Filer Features](#filer-features)
-* [Example Usage](#example-usage)
+* [Example: Using Seaweed Object Store](#example-Using-Seaweed-Object-Store)
 * [Architecture](#architecture)
 * [Compared to Other File Systems](#compared-to-other-file-systems)
    * [Compared to HDFS](#compared-to-hdfs)
@@ -75,6 +65,13 @@ Table of Contents
 * [Benchmark](#Benchmark)
 * [License](#license)

+
+## Quick Start ##
+* 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 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 ##

 SeaweedFS is a simple and highly scalable distributed file system. There are two objectives:
@@ -82,15 +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)

@@ -101,7 +100,7 @@ On top of the object store, optional [Filer] can support directories and POSIX a
 * Automatic compaction to reclaim disk space after deletion or update.
 * [Automatic entry TTL expiration][VolumeServerTTL].
 * Any server with some disk spaces can add to the total storage space.
-* Adding/Removing servers does **not** cause any data re-balancing.
+* Adding/Removing servers does **not** cause any data re-balancing unless triggered by admin commands.
 * Optional picture resizing.
 * Support ETag, Accept-Range, Last-Modified, etc.
 * Support in-memory/leveldb/readonly mode tuning for memory/performance balance.
@@ -113,16 +112,20 @@ On top of the object store, optional [Filer] can support directories and POSIX a

 ## Filer Features ##
 * [Filer server][Filer] provides "normal" directories and files via http.
-* [Super Large Files][SuperLargeFiles] stores large or super large files in tens of TB.
+* [File TTL][FilerTTL] automatically expires file metadata and actual file data.
 * [Mount filer][Mount] reads and writes files directly as a local directory via FUSE.
+* [Filer Store Replication][FilerStoreReplication] enables HA for filer meta data stores.
 * [Active-Active Replication][ActiveActiveAsyncReplication] enables asynchronous one-way or two-way cross cluster continuous replication.
 * [Amazon S3 compatible API][AmazonS3API] accesses files with S3 tooling.
 * [Hadoop Compatible File System][Hadoop] accesses files from Hadoop/Spark/Flink/etc or even runs HBase.
 * [Async Replication To Cloud][BackupToCloud] has extremely fast local access and backups to Amazon S3, Google Cloud Storage, Azure, BackBlaze.
 * [WebDAV] accesses as a mapped drive on Mac and Windows, or from mobile devices.
 * [AES256-GCM Encrypted Storage][FilerDataEncryption] safely stores the encrypted data.
-* [File TTL][FilerTTL] automatically purges file metadata and actual file 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
@@ -138,10 +141,12 @@ On top of the object store, optional [Filer] can support directories and POSIX a
 [VolumeServerTTL]: https://github.com/chrislusf/seaweedfs/wiki/Store-file-with-a-Time-To-Live
 [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)

-## Example Usage ##
+## Example: Using Seaweed Object Store ##

 By default, the master node runs on port 9333, and the volume nodes run on port 8080.
 Let's start one master node, and two volume nodes on port 8080 and 8081. Ideally, they should be started from different machines. We'll use localhost as an example.
@@ -339,7 +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)

@@ -369,7 +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 |
@@ -407,11 +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 |
 | -------------  | ------------- | ---------------- |
@@ -425,9 +430,10 @@ SeaweedFS Filer uses off-the-shelf stores, such as MySql, Postgres, Mongodb, Red

 MinIO follows AWS S3 closely and is ideal for testing for S3 API. It has good UI, policies, versionings, etc. SeaweedFS is trying to catch up here. It is also possible to put MinIO as a gateway in front of SeaweedFS later.

-MinIO metadata are in simple files. Each file write will incur meta file writes.
+MinIO metadata are in simple files. Each file write will incur extra writes to corresponding meta file.

-MinIO does not have optimization for large number of small files.
+MinIO does not have optimization for lots of small files. The files are simply stored as is to local disks.
+Plus the extra meta file and shards for erasure coding, it only amplifies the LOSF problem.

 MinIO has multiple disk IO to read one file. SeaweedFS has O(1) disk reads, even for erasure coded files.

@@ -439,17 +445,12 @@ MinIO has specific requirements on storage layout. It is not flexible to adjust

 ## Dev Plan ##

-More tools and documentation, on how to maintain and scale the system. For example, how to move volumes, automatically balancing data, how to grow volumes, how to check system status, etc.
-Other key features include: Erasure Encoding, JWT security.
+* More tools and documentation, on how to manage and scale the system.
+* Read and write stream data.
+* Support structured data.

 This is a super exciting project! And we need helpers and [support](https://www.patreon.com/seaweedfs)!

-BTW, We suggest run the code style check script `util/gostd` before you push your branch to remote, it will make SeaweedFS easy to review, maintain and develop:
-
-```
-$ ./util/gostd
-```
-
 [Back to TOC](#table-of-contents)

 ## Installation Guide ##
@@ -1,5 +1,7 @@
 FROM alpine

+ARG RELEASE=latest # 'latest' or 'dev'
+
 RUN \
    ARCH=$(if [ $(uname -m) == "x86_64" ] && [ $(getconf LONG_BIT) == "64" ]; then echo "amd64"; \
         elif [ $(uname -m) == "x86_64" ] && [ $(getconf LONG_BIT) == "32" ]; then echo "386"; \
@@ -12,7 +14,8 @@ RUN \
    SUPERCRONIC=supercronic-linux-$ARCH && \
    # Install SeaweedFS and Supercronic ( for cron job mode )
    apk add --no-cache --virtual build-dependencies --update wget curl ca-certificates && \
-    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") && \
+    apk add fuse && \
+    wget -P /tmp https://github.com/$(curl -s -L https://github.com/chrislusf/seaweedfs/releases/${RELEASE} | egrep -o "chrislusf/seaweedfs/releases/download/.*/linux_$ARCH.tar.gz") && \
    tar -C /usr/bin/ -xzvf /tmp/linux_$ARCH.tar.gz && \
    curl -fsSLO "$SUPERCRONIC_URL" && \
    echo "${SUPERCRONIC_SHA1SUM}  ${SUPERCRONIC}" | sha1sum -c - && \
@@ -36,6 +39,8 @@ EXPOSE 19333
 EXPOSE 9333
 # s3 server http port
 EXPOSE 8333
+# webdav server http port
+EXPOSE 7333

 RUN mkdir -p /data/filerldb2

@@ -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
@@ -27,6 +30,8 @@ EXPOSE 19333
 EXPOSE 9333
 # s3 server http port
 EXPOSE 8333
+# webdav server http port
+EXPOSE 7333

 RUN mkdir -p /data/filerldb2

@@ -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
@@ -27,6 +30,8 @@ EXPOSE 19333
 EXPOSE 9333
 # s3 server http port
 EXPOSE 8333
+# webdav server http port
+EXPOSE 7333

 RUN mkdir -p /data/filerldb2

@@ -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
@@ -19,6 +20,8 @@ EXPOSE 19333
 EXPOSE 9333
 # s3 server http port
 EXPOSE 8333
+# webdav server http port
+EXPOSE 7333

 RUN mkdir -p /data/filerldb2

@@ -0,0 +1,31 @@
+FROM ubuntu:20.04
+
+RUN DEBIAN_FRONTEND=noninteractive apt-get update && \
+    DEBIAN_FRONTEND=noninteractive apt-get upgrade -y && \
+    DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends \
+        git \
+        sudo \
+        debianutils \
+        python3-pip \
+        python3-virtualenv \
+        python3-dev \
+        libevent-dev \
+        libffi-dev \
+        libxml2-dev \
+        libxslt-dev \
+        zlib1g-dev && \
+    DEBIAN_FRONTEND=noninteractive apt-get clean && \
+    rm -rf /var/lib/apt/lists/* && \
+    git clone https://github.com/ceph/s3-tests.git /opt/s3-tests
+
+WORKDIR /opt/s3-tests
+RUN ./bootstrap
+
+ENV \
+    NOSETESTS_EXCLUDE="" \
+    NOSETESTS_ATTR="" \
+    NOSETESTS_OPTIONS="" \
+    S3TEST_CONF="/s3test.conf"
+
+ENTRYPOINT ["/bin/bash", "-c"]
+CMD ["exec ./virtualenv/bin/nosetests ${NOSETESTS_OPTIONS-} ${NOSETESTS_ATTR:+-a $NOSETESTS_ATTR} ${NOSETESTS_EXCLUDE:+-e $NOSETESTS_EXCLUDE}"]
@@ -9,11 +9,35 @@ build:
 	docker build --no-cache -t chrislusf/seaweedfs:local -f Dockerfile.local .
 	rm ./weed

+s3tests_build:
+	docker build --no-cache -t chrislusf/ceph-s3-tests:local -f Dockerfile.s3tests .
+
 dev: build
-	docker-compose -f local-dev-compose.yml -p seaweedfs up
+	docker-compose -f compose/local-dev-compose.yml -p seaweedfs up
+
+dev_mount: build
+	docker-compose -f compose/local-mount-compose.yml -p seaweedfs up
+
+k8s: build
+	docker-compose -f compose/local-k8s-compose.yml -p seaweedfs up
+
+dev_registry: build
+	docker-compose -f compose/local-registry-compose.yml -p seaweedfs up
+
+dev_replicate: build
+	docker-compose -f compose/local-replicate-compose.yml -p seaweedfs up

 cluster: build
-	docker-compose -f local-cluster-compose.yml -p seaweedfs up
+	docker-compose -f compose/local-cluster-compose.yml -p seaweedfs up
+
+2clusters: build
+	docker-compose -f compose/local-clusters-compose.yml -p seaweedfs up
+
+s3tests: build s3tests_build
+	docker-compose -f compose/local-s3tests-compose.yml -p seaweedfs up
+
+filer_etcd: build
+	docker stack deploy -c compose/swarm-etcd.yml fs

 clean:
 	rm ./weed
@@ -19,12 +19,32 @@ services:
      - 9335:9335
      - 19335:19335
    command: "master -ip=master2 -port=9335 -peers=master0:9333,master1:9334,master2:9335 -mdir=/data/m3"
-  volume:
+  volume1:
    image: chrislusf/seaweedfs:local
    ports:
      - 8080:8080
      - 18080:18080
-    command: 'volume -mserver="master0:9333,master1:9334,master2:9335" -port=8080 -ip=volume -publicUrl=localhost:8080'
+    command: 'volume -mserver="master0:9333,master1:9334,master2:9335" -port=8080 -ip=volume1 -publicUrl=localhost:8080 -preStopSeconds=1 -disk=ssd1'
+    depends_on:
+      - master0
+      - master1
+      - master2
+  volume2:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8082:8082
+      - 18082:18082
+    command: 'volume -mserver="master0:9333,master1:9334,master2:9335" -port=8082 -ip=volume2 -publicUrl=localhost:8082 -preStopSeconds=1 -disk=ssd1'
+    depends_on:
+      - master0
+      - master1
+      - master2
+  volume3:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8083:8083
+      - 18083:18083
+    command: 'volume -mserver="master0:9333,master1:9334,master2:9335" -port=8083 -ip=volume3 -publicUrl=localhost:8083 -preStopSeconds=1'
    depends_on:
      - master0
      - master1
@@ -39,7 +59,8 @@ services:
      - master0
      - master1
      - master2
-      - volume
+      - volume1
+      - volume2
  s3:
    image: chrislusf/seaweedfs:local
    ports:
@@ -49,5 +70,6 @@ services:
      - master0
      - master1
      - master2
-      - volume
+      - volume1
+      - volume2
      - filer
@@ -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
@@ -0,0 +1,91 @@
+version: '2'
+
+services:
+  master:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9333:9333
+      - 19333:19333
+    command: "master -ip=master -volumeSizeLimitMB=1024"
+  volume:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8080:8080
+      - 18080:18080
+    command: "volume -mserver=master:9333 -port=8080 -ip=volume -max=0 -preStopSeconds=1"
+    depends_on:
+      - master
+  filer:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8888:8888
+      - 18888:18888
+    command: 'filer -master="master:9333"'
+    depends_on:
+      - master
+      - volume
+  s3:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8333:8333
+    command: '-v 9 s3 -filer="filer:8888"'
+    depends_on:
+      - master
+      - volume
+      - filer
+  minio:
+    image: minio/minio
+    ports:
+      - 9000:9000
+    command: 'minio server /data'
+    environment:
+      MINIO_ACCESS_KEY: "some_access_key1"
+      MINIO_SECRET_KEY: "some_secret_key1"
+    depends_on:
+      - master
+  registry1:
+    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://s3:8333"
+      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
+      - minio
+  registry2:
+    image: registry:2
+    environment:
+      REGISTRY_HTTP_ADDR: "0.0.0.0:5002" # minio
+      REGISTRY_LOG_LEVEL: "debug"
+      REGISTRY_STORAGE: "s3"
+      REGISTRY_STORAGE_S3_REGION: "us-east-1"
+      REGISTRY_STORAGE_S3_REGIONENDPOINT: "http://minio:9000"
+      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:
+      - 5002:5002
+    depends_on:
+      - s3
+      - minio
@@ -0,0 +1,61 @@
+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
+    environment:
+      RABBITMQ_SERVER_ADDITIONAL_ERL_ARGS: "-rabbit log_levels [{connection,error},{queue,debug}]"
@@ -0,0 +1,45 @@
+version: '2'
+
+services:
+  master:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9333:9333
+      - 19333:19333
+    command: "master -ip=master -volumeSizeLimitMB=16"
+    environment:
+      WEED_MASTER_VOLUME_GROWTH_COPY_1: 1
+      WEED_MASTER_VOLUME_GROWTH_COPY_OTHER: 1
+  volume:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8080:8080
+      - 18080:18080
+    command: "volume -mserver=master:9333 -port=8080 -ip=volume -preStopSeconds=1"
+    depends_on:
+      - master
+  s3:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8888:8888
+      - 18888:18888
+      - 8000:8000
+    command: 'filer -master="master:9333" -s3 -s3.config=/etc/seaweedfs/s3.json -s3.port=8000'
+    volumes:
+      - ./s3.json:/etc/seaweedfs/s3.json
+    depends_on:
+      - master
+      - volume
+  s3tests:
+    image: chrislusf/ceph-s3-tests:local
+    volumes:
+      - ./s3tests.conf:/opt/s3-tests/s3tests.conf
+    environment:
+      S3TEST_CONF: "s3tests.conf"
+      NOSETESTS_OPTIONS: "--verbose --logging-level=ERROR --with-xunit --failure-detail s3tests_boto3.functional.test_s3"
+      NOSETESTS_ATTR: "!tagging,!fails_on_aws,!encryption,!bucket-policy,!versioning,!fails_on_rgw,!bucket-policy,!fails_with_subdomain,!policy_status,!object-lock,!lifecycle,!cors,!user-policy"
+      NOSETESTS_EXCLUDE: "(bucket_list_delimiter_basic|bucket_listv2_delimiter_basic|bucket_listv2_encoding_basic|bucket_list_encoding_basic|bucket_list_delimiter_prefix|bucket_listv2_delimiter_prefix_ends_with_delimiter|bucket_list_delimiter_prefix_ends_with_delimiter|bucket_list_delimiter_alt|bucket_listv2_delimiter_alt|bucket_list_delimiter_prefix_underscore|bucket_list_delimiter_percentage|bucket_listv2_delimiter_percentage|bucket_list_delimiter_whitespace|bucket_listv2_delimiter_whitespace|bucket_list_delimiter_dot|bucket_listv2_delimiter_dot|bucket_list_delimiter_unreadable|bucket_listv2_delimiter_unreadable|bucket_listv2_fetchowner_defaultempty|bucket_listv2_fetchowner_empty|bucket_list_delimiter_not_skip_special|bucket_list_prefix_delimiter_alt|bucket_listv2_prefix_delimiter_alt|bucket_list_prefix_delimiter_prefix_not_exist|bucket_listv2_prefix_delimiter_prefix_not_exist|bucket_list_prefix_delimiter_delimiter_not_exist|bucket_listv2_prefix_delimiter_delimiter_not_exist|bucket_list_prefix_delimiter_prefix_delimiter_not_exist|bucket_listv2_prefix_delimiter_prefix_delimiter_not_exist|bucket_list_maxkeys_none|bucket_listv2_maxkeys_none|bucket_list_maxkeys_invalid|bucket_listv2_continuationtoken_empty|bucket_list_return_data|bucket_list_objects_anonymous|bucket_listv2_objects_anonymous|bucket_notexist|bucketv2_notexist|bucket_delete_nonempty|bucket_concurrent_set_canned_acl|object_write_to_nonexist_bucket|object_requestid_matches_header_on_error|object_head_zero_bytes|object_write_cache_control|object_write_expires|object_set_get_metadata_none_to_good|object_set_get_metadata_none_to_empty|object_set_get_metadata_overwrite_to_empty|post_object_anonymous_request|post_object_authenticated_request|post_object_authenticated_no_content_type|post_object_authenticated_request_bad_access_key|post_object_set_success_code|post_object_set_invalid_success_code|post_object_upload_larger_than_chunk|post_object_set_key_from_filename|post_object_ignored_header|post_object_case_insensitive_condition_fields|post_object_escaped_field_values|post_object_success_redirect_action|post_object_invalid_signature|post_object_invalid_access_key|post_object_missing_policy_condition|post_object_user_specified_header|post_object_request_missing_policy_specified_field|post_object_expired_policy|post_object_invalid_request_field_value|get_object_ifmatch_failed|get_object_ifunmodifiedsince_good|put_object_ifmatch_failed|object_raw_get|object_raw_get_bucket_gone|object_delete_key_bucket_gone|object_raw_get_bucket_acl|object_raw_get_object_acl|object_raw_authenticated|object_raw_response_headers|object_raw_authenticated_bucket_acl|object_raw_authenticated_object_acl|object_raw_authenticated_bucket_gone|object_raw_get_x_amz_expires_not_expired|object_raw_get_x_amz_expires_out_max_range|object_raw_get_x_amz_expires_out_positive_range|object_anon_put_write_access|object_raw_put_authenticated_expired|bucket_create_naming_bad_short_one|bucket_create_naming_bad_short_two|bucket_create_exists|bucket_get_location|bucket_acl_default|bucket_acl_canned|bucket_acl_canned_publicreadwrite|bucket_acl_canned_authenticatedread|object_acl_default|object_acl_canned_during_create|object_acl_canned|object_acl_canned_publicreadwrite|object_acl_canned_authenticatedread|object_acl_canned_bucketownerread|object_acl_canned_bucketownerfullcontrol|object_acl_full_control_verify_attributes|bucket_acl_canned_private_to_private|bucket_acl_grant_nonexist_user|bucket_acl_no_grants|bucket_acl_grant_email_not_exist|bucket_acl_revoke_all|bucket_recreate_not_overriding|bucket_create_special_key_names|object_copy_zero_size|object_copy_verify_contenttype|object_copy_to_itself|object_copy_to_itself_with_metadata|object_copy_not_owned_bucket|object_copy_not_owned_object_bucket|object_copy_retaining_metadata|object_copy_replacing_metadata|multipart_upload_empty|multipart_copy_invalid_range|multipart_copy_special_names|multipart_upload_resend_part|multipart_upload_size_too_small|abort_multipart_upload_not_found|multipart_upload_missing_part|multipart_upload_incorrect_etag|100_continue|ranged_request_invalid_range|ranged_request_empty_object|access_bucket)"
+    depends_on:
+      - master
+      - volume
+      - s3
@@ -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 +1,11 @@
+[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"
@@ -0,0 +1,105 @@
+{
+  "identities": [
+    {
+      "name": "anonymous",
+      "actions": [
+        "Read"
+      ]
+    },
+    {
+      "name": "some_admin_user",
+      "credentials": [
+        {
+          "accessKey": "some_access_key1",
+          "secretKey": "some_secret_key1"
+        }
+      ],
+      "actions": [
+        "Admin",
+        "Read",
+        "List",
+        "Tagging",
+        "Write"
+      ]
+    },
+    {
+      "name": "s3_tests",
+      "credentials": [
+        {
+          "accessKey": "ABCDEFGHIJKLMNOPQRST",
+          "secretKey": "abcdefghijklmnopqrstuvwxyzabcdefghijklmn"
+        },
+        {
+          "accessKey": "0555b35654ad1656d804",
+          "secretKey": "h7GhxuBLTrlhVUyxSPUKUV8r/2EI4ngqJxD7iBdBYLhwluN30JaT3Q=="
+        }
+      ],
+      "actions": [
+        "Admin",
+        "Read",
+        "List",
+        "Tagging",
+        "Write"
+      ]
+    },
+    {
+      "name": "s3_tests_alt",
+      "credentials": [
+        {
+          "accessKey": "NOPQRSTUVWXYZABCDEFG",
+          "secretKey": "nopqrstuvwxyzabcdefghijklmnabcdefghijklm"
+        }
+      ],
+      "actions": [
+        "Admin",
+        "Read",
+        "List",
+        "Tagging",
+        "Write"
+      ]
+    },
+    {
+      "name": "s3_tests_tenant",
+      "credentials": [
+        {
+          "accessKey": "HIJKLMNOPQRSTUVWXYZA",
+          "secretKey": "opqrstuvwxyzabcdefghijklmnopqrstuvwxyzab"
+        }
+      ],
+      "actions": [
+        "Admin",
+        "Read",
+        "List",
+        "Tagging",
+        "Write"
+      ]
+    },
+    {
+      "name": "some_read_only_user",
+      "credentials": [
+        {
+          "accessKey": "some_access_key2",
+          "secretKey": "some_secret_key2"
+        }
+      ],
+      "actions": [
+        "Read"
+      ]
+    },
+    {
+      "name": "some_normal_user",
+      "credentials": [
+        {
+          "accessKey": "some_access_key3",
+          "secretKey": "some_secret_key3"
+        }
+      ],
+      "actions": [
+        "Read",
+        "List",
+        "Tagging",
+        "Write"
+      ]
+    }
+  ]
+}
@@ -0,0 +1,70 @@
+[DEFAULT]
+## this section is just used for host, port and bucket_prefix
+
+# host set for rgw in vstart.sh
+host = s3
+
+# port set for rgw in vstart.sh
+port = 8000
+
+## say "False" to disable TLS
+is_secure = False
+
+[fixtures]
+## all the buckets created will start with this prefix;
+## {random} will be filled with random characters to pad
+## the prefix to 30 characters long, and avoid collisions
+bucket prefix = yournamehere-{random}-
+
+[s3 main]
+# main display_name set in vstart.sh
+display_name = M. Tester
+
+# main user_idname set in vstart.sh
+user_id = testid
+
+# main email set in vstart.sh
+email = tester@ceph.com
+
+# zonegroup api_name for bucket location
+api_name = default
+
+## main AWS access key
+access_key = 0555b35654ad1656d804
+
+## main AWS secret key
+secret_key = h7GhxuBLTrlhVUyxSPUKUV8r/2EI4ngqJxD7iBdBYLhwluN30JaT3Q==
+
+## replace with key id obtained when secret is created, or delete if KMS not tested
+#kms_keyid = 01234567-89ab-cdef-0123-456789abcdef
+
+[s3 alt]
+# alt display_name set in vstart.sh
+display_name = john.doe
+## alt email set in vstart.sh
+email = john.doe@example.com
+
+# alt user_id set in vstart.sh
+user_id = 56789abcdef0123456789abcdef0123456789abcdef0123456789abcdef01234
+
+# alt AWS access key set in vstart.sh
+access_key = NOPQRSTUVWXYZABCDEFG
+
+# alt AWS secret key set in vstart.sh
+secret_key = nopqrstuvwxyzabcdefghijklmnabcdefghijklm
+
+[s3 tenant]
+# tenant display_name set in vstart.sh
+display_name = testx$tenanteduser
+
+# tenant user_id set in vstart.sh
+user_id = 9876543210abcdef0123456789abcdef0123456789abcdef0123456789abcdef
+
+# tenant AWS secret key set in vstart.sh
+access_key = HIJKLMNOPQRSTUVWXYZA
+
+# tenant AWS secret key set in vstart.sh
+secret_key = opqrstuvwxyzabcdefghijklmnopqrstuvwxyzab
+
+# tenant email set in vstart.sh
+email = tenanteduser@example.com
@@ -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:
@@ -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 +1,13 @@
+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'
@@ -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 +42,27 @@ 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
+  webdav:
+    image: chrislusf/seaweedfs # use a remote image
+    ports:
+      - 7333:7333
+    command: 'webdav -filer="filer:8888"'
+    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
@@ -33,3 +33,12 @@ services:
      - master
      - volume
      - filer
+  webdav:
+    image: chrislusf/seaweedfs:dev # use a remote dev image
+    ports:
+      - 7333:7333
+    command: 'webdav -filer="filer:8888"'
+    depends_on:
+      - master
+      - volume
+      - filer
@@ -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;
@@ -3,15 +3,18 @@ module github.com/chrislusf/seaweedfs
 go 1.12

 require (
-	cloud.google.com/go v0.44.3
+	cloud.google.com/go v0.58.0 // indirect
+	cloud.google.com/go/pubsub v1.3.1
+	cloud.google.com/go/storage v1.9.0
+	github.com/Azure/azure-amqp-common-go/v2 v2.1.0 // indirect
 	github.com/Azure/azure-pipeline-go v0.2.2 // indirect
-	github.com/Azure/azure-storage-blob-go v0.8.0
+	github.com/Azure/azure-storage-blob-go v0.9.0
 	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 v1.0.2-0.20201002174524-b13c3bfdb011
+	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
@@ -22,12 +25,14 @@ require (
 	github.com/facebookgo/stack v0.0.0-20160209184415-751773369052 // indirect
 	github.com/facebookgo/stats v0.0.0-20151006221625-1b76add642e4
 	github.com/facebookgo/subset v0.0.0-20200203212716-c811ad88dec4 // indirect
+	github.com/fclairamb/ftpserverlib v0.8.0
 	github.com/frankban/quicktest v1.7.2 // indirect
-	github.com/go-redis/redis v6.15.7+incompatible
+	github.com/go-errors/errors v1.1.1 // indirect
+	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
-	github.com/golang/groupcache v0.0.0-20190702054246-869f871628b6
+	github.com/golang/groupcache v0.0.0-20200121045136-8c9f03a8e57e
 	github.com/golang/protobuf v1.4.2
 	github.com/google/btree v1.0.0
 	github.com/google/uuid v1.1.1
@@ -36,55 +41,56 @@ require (
 	github.com/grpc-ecosystem/grpc-gateway v1.11.0 // indirect
 	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/klauspost/compress v1.10.9
+	github.com/karlseguin/ccache v2.0.3+incompatible // 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
 	github.com/klauspost/reedsolomon v1.9.2
 	github.com/kurin/blazer v0.5.3
 	github.com/lib/pq v1.2.0
+	github.com/lunixbochs/vtclean v1.0.0 // indirect
 	github.com/magiconair/properties v1.8.1 // indirect
-	github.com/mattn/go-ieproxy v0.0.0-20190805055040-f9202b1cfdeb // indirect
+	github.com/mattn/go-colorable v0.1.2 // indirect
 	github.com/mattn/go-runewidth v0.0.4 // indirect
-	github.com/nats-io/nats-server/v2 v2.0.4 // 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 v1.1.0
-	github.com/prometheus/procfs v0.0.4 // 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
-	github.com/spf13/afero v1.2.2 // indirect
 	github.com/spf13/jwalterweatherman v1.1.0 // indirect
 	github.com/spf13/viper v1.4.0
-	github.com/streadway/amqp v0.0.0-20190827072141-edfb9018d271 // indirect
+	github.com/streadway/amqp v0.0.0-20200108173154-1c71cc93ed71
 	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/valyala/fasthttp v1.20.0
+	github.com/viant/assertly v0.5.4 // indirect
+	github.com/viant/ptrie v0.3.0
+	github.com/viant/toolbox v0.33.2 // indirect
 	github.com/willf/bitset v1.1.10 // indirect
 	github.com/willf/bloom v2.0.3+incompatible
 	github.com/wsxiaoys/terminal v0.0.0-20160513160801-0940f3fc43a0 // indirect
 	go.etcd.io/etcd v3.3.15+incompatible
 	go.mongodb.org/mongo-driver v1.3.2
-	gocloud.dev v0.16.0
-	gocloud.dev/pubsub/natspubsub v0.16.0
-	gocloud.dev/pubsub/rabbitpubsub v0.16.0
+	gocloud.dev v0.20.0
+	gocloud.dev/pubsub/natspubsub v0.20.0
+	gocloud.dev/pubsub/rabbitpubsub v0.20.0
 	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 v0.0.0-20191029190741-b9c20aec41a5
-	google.golang.org/api v0.9.0
-	google.golang.org/appengine v1.6.2 // indirect
+	golang.org/x/tools v0.0.0-20200608174601-1b747fd94509
+	google.golang.org/api v0.26.0
 	google.golang.org/grpc v1.29.1
 	google.golang.org/protobuf v1.24.0
 	gopkg.in/check.v1 v1.0.0-20190902080502-41f04d3bba15 // indirect
@@ -9,15 +9,32 @@ and backup/HA memsql can provide.
 with ENV.
 * cert config exists and can be enabled, but not been tested.

-### current instances config (AIO):
-1 instance for each type (master/filer/volume/s3)
+### prerequisites
+kubernetes node have labels which help to define which node(Host) will run which pod.

-instances need node labels:
-* sw-volume: true  (for volume instance, specific tag)
-* sw-backend: true (for all others, as they less resource demanding)
+s3/filer/master needs the label **sw-backend=true**
+
+volume need the label **sw-volume=true**
+
+to label a node to be able to run all pod types in k8s:
+```
+kubectl label node YOUR_NODE_NAME sw-volume=true,sw-backend=true
+```
+
+on production k8s deployment you will want each pod to have a different host,
+especially the volume server & the masters, currently all pods (master/volume/filer)
+have anti-affinity rule to disallow running multiple pod type on the same host.
+if you still want to run multiple pods of the same type (master/volume/filer) on the same host
+please set/update the corresponding affinity rule in values.yaml to an empty one:
+
+```affinity: ""```
+
+
+### current instances config (AIO):
+1 instance for each type (master/filer+s3/volume)

 you can update the replicas count for each node type in values.yaml,
-need to add more nodes with the corresponding label.
+need to add more nodes with the corresponding labels.

 most of the configuration are available through values.yaml

@@ -1,4 +1,5 @@
 apiVersion: v1
 description: SeaweedFS
 name: seaweedfs
-version: 2.06
+appVersion: "2.27"
+version: 2.27
@@ -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
@@ -14,13 +15,13 @@ spec:
      backoffLimit: 2
      template:
        spec:
-          {{- with .Values.cronjob.nodeSelector }}
+          {{- if .Values.cronjob.nodeSelector }}
          nodeSelector:
-          {{- toYaml . | nindent 12 }}
+            {{ tpl .Values.cronjob.nodeSelector . | indent 12 | trim }}
          {{- end }}
-          {{- with .Values.cronjob.tolerations }}
+          {{- if .Values.cronjob.tolerations }}
          tolerations:
-          {{- toYaml . | nindent 12 }}
+            {{ tpl .Values.cronjob.tolerations . | nindent 12 | trim }}
          {{- end }}
          restartPolicy: OnFailure
          containers:
@@ -35,10 +36,13 @@ spec:
                - |
                  set -ex
                  echo -e "lock\n\
-                  volume.balance -force\
+                  volume.balance -force \
                  {{ if .Values.volume.dataCenter }} -dataCenter {{ .Values.volume.dataCenter }}{{ end }}\
                  {{ if .Values.cronjob.collection }} -collection {{ .Values.cronjob.collection }}{{ end }}\n\
-                  volume.fix.replication\nunlock\n" | \
+                  {{- if .Values.cronjob.enableFixReplication }}
+                  volume.fix.replication {{ if .Values.cronjob.collectionPattern }} -collectionPattern={{ .Values.cronjob.collectionPattern }} {{ end }} \n\
+                  {{- end }}
+                  unlock\n" | \
                  /usr/bin/weed shell \
                  {{- if .Values.cronjob.master }}
                  -master {{ .Values.cronjob.master }} \
@@ -51,4 +55,4 @@ spec:
                  -filer {{ template "seaweedfs.name" . }}-filer.{{ .Release.Namespace }}.svc:{{ .Values.filer.port }}
                  {{- end }}
 {{- end }}
-{{- end }}
+{{- end }}
@@ -0,0 +1,30 @@
+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:
+  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: {{ $value | quote }}
+            {{- end }}
+            {{- end }}
          command:
            - "/bin/sh"
            - "-ec"
@@ -101,7 +107,7 @@ spec:
              -port={{ .Values.filer.port }} \
              {{- if .Values.filer.metricsPort }}
              -metricsPort {{ .Values.filer.metricsPort }} \
-              {{- end }}}
+              {{- end }}
              {{- if .Values.filer.redirectOnRead }}
              -redirectOnRead \
              {{- end }}
@@ -127,14 +133,36 @@ spec:
              -encryptVolumeData \
              {{- end }}
              -ip=${POD_IP} \
+              {{- if .Values.filer.enable_peers }}
              {{- 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 }}
+              {{- end }}
+              {{- if .Values.filer.s3.enabled }}
+              -s3 \
+              -s3.port={{ .Values.filer.s3.port }} \
+              {{- if .Values.filer.s3.domainName }}
+              -s3.domainName={{ .Values.filer.s3.domainName }} \
+              {{- end }}
+              {{- if .Values.global.enableSecurity }}
+              -s3.cert.file=/usr/local/share/ca-certificates/client/tls.crt \
+              -s3.key.file=/usr/local/share/ca-certificates/client/tls.key \
+              {{- end }}
+              {{- if .Values.filer.s3.allowEmptyFolder }}
+              -s3.allowEmptyFolder={{ .Values.filer.s3.allowEmptyFolder }} \
+              {{- end }}
+              {{- if .Values.filer.s3.enableAuth }}
+              -s3.config=/etc/sw/seaweedfs_s3_config \
+              {{- end }}
              {{- 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) }}    
          volumeMounts:
            - name: seaweedfs-filer-log-volume
              mountPath: "/logs/"
+            - mountPath: /etc/sw
+              name: config-users
+              readOnly: true
            {{- if .Values.global.enableSecurity }}
            - name: security-config
              readOnly: true
@@ -172,7 +200,7 @@ spec:
            periodSeconds: 15
            successThreshold: 1
            failureThreshold: 100
-            timeoutSeconds: 3
+            timeoutSeconds: 10
          livenessProbe:
            httpGet:
              path: /
@@ -182,7 +210,7 @@ spec:
            periodSeconds: 30
            successThreshold: 1
            failureThreshold: 5
-            timeoutSeconds: 3
+            timeoutSeconds: 10
          {{- if .Values.filer.resources }}
          resources:
            {{ tpl .Values.filer.resources . | nindent 12 | trim }}
@@ -192,6 +220,13 @@ spec:
          hostPath:
            path: /storage/logs/seaweedfs/filer
            type: DirectoryOrCreate
+        - name: db-schema-config-volume
+          configMap:
+            name: seaweedfs-db-init-config
+        - name: config-users
+          secret:
+            defaultMode: 420
+            secretName: seaweedfs-s3-secret
        {{- if .Values.global.enableSecurity }}
        - name: security-config
          configMap:
@@ -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 }}
@@ -73,7 +79,7 @@ spec:
              -port={{ .Values.s3.port }} \
              {{- if .Values.s3.metricsPort }}
              -metricsPort {{ .Values.s3.metricsPort }} \
-              {{- end }}}
+              {{- end }}
              {{- if .Values.global.enableSecurity }}
              -cert.file=/usr/local/share/ca-certificates/client/tls.crt \
              -key.file=/usr/local/share/ca-certificates/client/tls.key \
@@ -81,9 +87,19 @@ 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 }}
+              {{- if .Values.s3.enableAuth }}
+              -config=/etc/sw/seaweedfs_s3_config \
+              {{- end }}
+              -filer={{ template "seaweedfs.name" . }}-filer-client:{{ .Values.filer.port }}
          volumeMounts:
+            - name: logs
+              mountPath: "/logs/"
+            - mountPath: /etc/sw
+              name: config-users
+              readOnly: true
            {{- if .Values.global.enableSecurity }}
            - name: security-config
              readOnly: true
@@ -106,7 +122,6 @@ 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 +134,7 @@ spec:
            periodSeconds: 15
            successThreshold: 1
            failureThreshold: 100
-            timeoutSeconds: 3
+            timeoutSeconds: 10
          livenessProbe:
            httpGet:
              path: /
@@ -129,12 +144,22 @@ 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:
+        - name: config-users
+          secret:
+            defaultMode: 420
+            secretName: seaweedfs-s3-secret
+        {{- 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:
@@ -9,15 +9,15 @@ metadata:
 spec:
  ports:
  - name: "swfs-s3"
-    port: {{ .Values.s3.port }}
-    targetPort: {{ .Values.s3.port }}
+    port: {{ if .Values.s3.enabled }}{{ .Values.s3.port }}{{ else }}{{ .Values.filer.s3.port }}{{ end }}
+    targetPort: {{ if .Values.s3.enabled }}{{ .Values.s3.port }}{{ else }}{{ .Values.filer.s3.port }}{{ end }}
    protocol: TCP
-{{- if .Values.s3.metricsPort }}
-  - name: "swfs-s3-metrics"
-    port: {{ .Values.filer.s3 }}
+{{- if and .Values.s3.enabled .Values.s3.metricsPort }}
+  - name: "metrics"
+    port: {{ .Values.s3.metricsPort }}
    targetPort: {{ .Values.s3.metricsPort }}
    protocol: TCP
-{{- end }}}
+{{- end }}
  selector:
    app: {{ template "seaweedfs.name" . }}
-    component: s3
+    component: {{ if .Values.s3.enabled }}s3{{ else }}filer{{ end }}
@@ -15,4 +15,4 @@ spec:
  selector:
    app: {{ template "seaweedfs.name" . }}
    component: s3
-{{- end }}}
+{{- end }}
@@ -0,0 +1,20 @@
+{{- if .Values.global.monitoring.enabled }}
+{{- $files := .Files.Glob "dashboards/*.json" }}
+{{- if $files }}
+apiVersion: v1
+kind: ConfigMapList
+items:
+{{- range $path, $fileContents := $files }}
+{{- $dashboardName := regexReplaceAll "(^.*/)(.*)\\.json$" $path "${2}" }}
+- apiVersion: v1
+  kind: ConfigMap
+  metadata:
+    name: {{ printf "%s" $dashboardName | lower | replace "_" "-" }}
+    namespace: {{ $.Release.Namespace }}
+    labels:
+     grafana_dashboard: "1"
+  data:
+    {{ $dashboardName }}.json: {{ $.Files.Get $path | toJson }}
+{{- end }}
+{{- end }}
+{{- end }}
@@ -0,0 +1,21 @@
+{{- if not (or .Values.filer.s3.skipAuthSecretCreation .Values.s3.skipAuthSecretCreation) }}
+{{- $access_key_admin := randAlphaNum 16 -}}
+{{- $secret_key_admin := randAlphaNum 32 -}}
+{{- $access_key_read := randAlphaNum 16 -}}
+{{- $secret_key_read := randAlphaNum 32 -}}
+apiVersion: v1
+kind: Secret
+type: Opaque
+metadata:
+  name: seaweedfs-s3-secret
+  namespace: {{ .Release.Namespace }}
+  annotations:
+    "helm.sh/resource-policy": keep
+    "helm.sh/hook": "pre-install"
+stringData:
+  admin_access_key_id: {{ $access_key_admin }}
+  admin_secret_access_key: {{ $secret_key_admin }}
+  read_access_key_id: {{ $access_key_read }}
+  read_secret_access_key: {{ $secret_key_read }}
+  seaweedfs_s3_config: '{"identities":[{"name":"anvAdmin","credentials":[{"accessKey":"{{ $access_key_admin }}","secretKey":"{{ $secret_key_admin }}"}],"actions":["Admin","Read","Write"]},{"name":"anvReadOnly","credentials":[{"accessKey":"{{ $access_key_read }}","secretKey":"{{ $secret_key_read }}"}],"actions":["Read"]}]}'
+{{- end }}
@@ -22,7 +22,7 @@ spec:
    port: {{ .Values.volume.metricsPort }}
    targetPort: {{ .Values.volume.metricsPort }}
    protocol: TCP
-{{- end }}}
+{{- end }}
 selector:
    app: {{ template "seaweedfs.name" . }}
    component: volume
@@ -15,4 +15,4 @@ spec:
  selector:
    app: {{ template "seaweedfs.name" . }}
    component: volume
-{{- end }}}
+{{- end }}
@@ -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"
@@ -78,8 +84,11 @@ spec:
                -port={{ .Values.volume.port }} \
                {{- if .Values.volume.metricsPort }}
                -metricsPort {{ .Values.volume.metricsPort }} \
-                {{- end }}}
+                {{- 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: "2.06"
+  # imageTag: "2.27" - 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: ""
@@ -132,7 +136,7 @@ volume:
  # limit file size to avoid out of memory, default 256mb
  fileSizeLimitMB: null
  # minimum free disk space(in percents). If free disk space lower this value - all volumes marks as ReadOnly
-  minFreeSpacePercent: 1
+  minFreeSpacePercent: 7


  # limit background compaction or copying speed in mega bytes per second
@@ -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")
@@ -223,6 +229,8 @@ filer:
  maxMB: null
  # encrypt data on volume servers
  encryptVolumeData: false
+  # enable peers sync metadata, for leveldb (localdb for filer but with sync across)
+  enable_peers: false

  # Whether proxy or redirect to volume server during file GET request
  redirectOnRead: false
@@ -292,8 +300,10 @@ 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"
@@ -302,13 +312,20 @@ filer:
    WEED_FILER_OPTIONS_RECURSIVE_DELETE: "false"
    # directories under this folder will be automatically creating a separate bucket
    WEED_FILER_BUCKETS_FOLDER: "/buckets"
-    # directories under this folder will be store message queue data
-    WEED_FILER_QUEUES_FOLDER: "/queues"
-    # WEED_FILER_OPTIONS_BUCKETS_FSYNC a list of buckets names with all write requests fsync=true
-    WEED_FILER_OPTIONS_BUCKETS_FSYNC: []
+
+  s3:
+    enabled: true
+    port: 8333
+    #allow empty folders
+    allowEmptyFolder: false
+    # Suffix of the host name, {bucket}.{domainName}
+    domainName: ""
+    # enable user & permission to s3 (need to inject to all services)
+    enableAuth: false
+    skipAuthSecretCreation: false

 s3:
-  enabled: true
+  enabled: false
  repository: null
  imageName: null
  imageTag: null
@@ -317,6 +334,11 @@ s3:
  port: 8333
  metricsPort: 9327
  loggingOverrideLevel: null
+  #allow empty folders
+  allowEmptyFolder: true
+  # enable user & permission to s3 (need to inject to all services)
+  enableAuth: false
+  skipAuthSecretCreation: false

  # Suffix of the host name, {bucket}.{domainName}
  domainName: ""
@@ -347,18 +369,27 @@ s3:
  # ref: https://kubernetes.io/docs/concepts/configuration/pod-priority-preemption/
  priorityClassName: ""

+  logs:
+    type: "hostPath"
+    size: ""
+    storageClass: ""
+
 cronjob:
-  enabled: false
+  enabled: true
+  master: "seaweedfs-master:9333"
+  filer: "seaweedfs-filer-client:8888"
+  tolerations: ""
+  nodeSelector: |
+    sw-backend: "true"
+  replication:
+    enable: true
+    collectionPattern: ""
  schedule: "*/7 * * * *"
  resources: null
  # balance all volumes among volume servers
  # ALL|EACH_COLLECTION|<collection_name>
  collection: ""
-  master: ""
-  filer: ""
-  tolerations: ""
-  nodeSelector: |
-    sw-backend: "true"
+

 certificates:
  commonName: "SeaweedFS CA"
@@ -5,7 +5,7 @@

    <groupId>com.github.chrislusf</groupId>
    <artifactId>seaweedfs-client</artifactId>
-    <version>1.5.1</version>
+    <version>1.6.1</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>

    <distributionManagement>
@@ -5,7 +5,7 @@

    <groupId>com.github.chrislusf</groupId>
    <artifactId>seaweedfs-client</artifactId>
-    <version>1.5.1</version>
+    <version>1.6.1</version>

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

    <groupId>com.github.chrislusf</groupId>
    <artifactId>seaweedfs-client</artifactId>
-    <version>1.5.1</version>
+    <version>1.6.1</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>
@@ -15,7 +15,6 @@ public class ChunkCache {
        }
        this.cache = CacheBuilder.newBuilder()
                .maximumSize(maxEntries)
-                .weakValues()
                .expireAfterAccess(1, TimeUnit.HOURS)
                .build();
    }
@@ -23,7 +23,7 @@ public class FileChunkManifest {
    }

    public static List<FilerProto.FileChunk> resolveChunkManifest(
-            final FilerGrpcClient filerGrpcClient, List<FilerProto.FileChunk> chunks) throws IOException {
+            final FilerClient filerClient, List<FilerProto.FileChunk> chunks) throws IOException {

        List<FilerProto.FileChunk> dataChunks = new ArrayList<>();

@@ -35,30 +35,30 @@ public class FileChunkManifest {

            // IsChunkManifest
            LOG.debug("fetching chunk manifest:{}", chunk);
-            byte[] data = fetchChunk(filerGrpcClient, chunk);
+            byte[] data = fetchChunk(filerClient, chunk);
            FilerProto.FileChunkManifest m = FilerProto.FileChunkManifest.newBuilder().mergeFrom(data).build();
            List<FilerProto.FileChunk> resolvedChunks = new ArrayList<>();
            for (FilerProto.FileChunk t : m.getChunksList()) {
                // avoid deprecated chunk.getFileId()
                resolvedChunks.add(t.toBuilder().setFileId(FilerClient.toFileId(t.getFid())).build());
            }
-            dataChunks.addAll(resolveChunkManifest(filerGrpcClient, resolvedChunks));
+            dataChunks.addAll(resolveChunkManifest(filerClient, resolvedChunks));
        }

        return dataChunks;
    }

-    private static byte[] fetchChunk(final FilerGrpcClient filerGrpcClient, FilerProto.FileChunk chunk) throws IOException {
+    private static byte[] fetchChunk(final FilerClient filerClient, FilerProto.FileChunk chunk) throws IOException {

        String vid = "" + chunk.getFid().getVolumeId();
-        FilerProto.Locations locations = filerGrpcClient.vidLocations.get(vid);
+        FilerProto.Locations locations = filerClient.vidLocations.get(vid);
        if (locations == null) {
            FilerProto.LookupVolumeRequest.Builder lookupRequest = FilerProto.LookupVolumeRequest.newBuilder();
            lookupRequest.addVolumeIds(vid);
-            FilerProto.LookupVolumeResponse lookupResponse = filerGrpcClient
+            FilerProto.LookupVolumeResponse lookupResponse = filerClient
                    .getBlockingStub().lookupVolume(lookupRequest.build());
            locations = lookupResponse.getLocationsMapMap().get(vid);
-            filerGrpcClient.vidLocations.put(vid, locations);
+            filerClient.vidLocations.put(vid, locations);
            LOG.debug("fetchChunk vid:{} locations:{}", vid, locations);
        }

@@ -74,7 +74,7 @@ public class FileChunkManifest {
        byte[] chunkData = SeaweedRead.chunkCache.getChunk(chunkView.fileId);
        if (chunkData == null) {
            LOG.debug("doFetchFullChunkData:{}", chunkView);
-            chunkData = SeaweedRead.doFetchFullChunkData(chunkView, locations);
+            chunkData = SeaweedRead.doFetchFullChunkData(filerClient, chunkView, locations);
        }
        if (chunk.getIsChunkManifest()){
            LOG.debug("chunk {} size {}", chunkView.fileId, chunkData.length);
@@ -86,7 +86,7 @@ public class FileChunkManifest {
    }

    public static List<FilerProto.FileChunk> maybeManifestize(
-            final FilerGrpcClient filerGrpcClient, List<FilerProto.FileChunk> inputChunks) throws IOException {
+            final FilerClient filerClient, List<FilerProto.FileChunk> inputChunks, String parentDirectory) throws IOException {
        // the return variable
        List<FilerProto.FileChunk> chunks = new ArrayList<>();

@@ -101,7 +101,7 @@ public class FileChunkManifest {

        int remaining = dataChunks.size();
        for (int i = 0; i + mergeFactor < dataChunks.size(); i += mergeFactor) {
-            FilerProto.FileChunk chunk = mergeIntoManifest(filerGrpcClient, dataChunks.subList(i, i + mergeFactor));
+            FilerProto.FileChunk chunk = mergeIntoManifest(filerClient, dataChunks.subList(i, i + mergeFactor), parentDirectory);
            chunks.add(chunk);
            remaining -= mergeFactor;
        }
@@ -113,7 +113,7 @@ public class FileChunkManifest {
        return chunks;
    }

-    private static FilerProto.FileChunk mergeIntoManifest(final FilerGrpcClient filerGrpcClient, List<FilerProto.FileChunk> dataChunks) throws IOException {
+    private static FilerProto.FileChunk mergeIntoManifest(final FilerClient filerClient, List<FilerProto.FileChunk> dataChunks, String parentDirectory) throws IOException {
        // create and serialize the manifest
        dataChunks = FilerClient.beforeEntrySerialization(dataChunks);
        FilerProto.FileChunkManifest.Builder m = FilerProto.FileChunkManifest.newBuilder().addAllChunks(dataChunks);
@@ -127,10 +127,10 @@ public class FileChunkManifest {
        }

        FilerProto.FileChunk.Builder manifestChunk = SeaweedWrite.writeChunk(
-                filerGrpcClient.getReplication(),
-                filerGrpcClient,
+                filerClient.getReplication(),
+                filerClient,
                minOffset,
-                data, 0, data.length);
+                data, 0, data.length, parentDirectory);
        manifestChunk.setIsChunkManifest(true);
        manifestChunk.setSize(maxOffset - minOffset);
        return manifestChunk.build();
@@ -4,25 +4,18 @@ import com.google.common.base.Strings;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;

-import java.nio.file.Path;
-import java.nio.file.Paths;
+import java.io.File;
 import java.util.ArrayList;
 import java.util.Arrays;
 import java.util.Iterator;
 import java.util.List;

-public class FilerClient {
+public class FilerClient extends FilerGrpcClient {

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

-    private final FilerGrpcClient filerGrpcClient;
-
    public FilerClient(String host, int grpcPort) {
-        filerGrpcClient = new FilerGrpcClient(host, grpcPort);
-    }
-
-    public FilerClient(FilerGrpcClient filerGrpcClient) {
-        this.filerGrpcClient = filerGrpcClient;
+        super(host, grpcPort);
    }

    public static String toFileId(FilerProto.FileId fid) {
@@ -100,9 +93,9 @@ public class FilerClient {
        if ("/".equals(path)) {
            return true;
        }
-        Path pathObject = Paths.get(path);
-        String parent = pathObject.getParent().toString();
-        String name = pathObject.getFileName().toString();
+        File pathFile = new File(path);
+        String parent = pathFile.getParent();
+        String name = pathFile.getName();

        mkdirs(parent, mode, uid, gid, userName, groupNames);

@@ -121,13 +114,13 @@ public class FilerClient {

    public boolean mv(String oldPath, String newPath) {

-        Path oldPathObject = Paths.get(oldPath);
-        String oldParent = oldPathObject.getParent().toString();
-        String oldName = oldPathObject.getFileName().toString();
+        File oldPathFile = new File(oldPath);
+        String oldParent = oldPathFile.getParent();
+        String oldName = oldPathFile.getName();

-        Path newPathObject = Paths.get(newPath);
-        String newParent = newPathObject.getParent().toString();
-        String newName = newPathObject.getFileName().toString();
+        File newPathFile = new File(newPath);
+        String newParent = newPathFile.getParent();
+        String newName = newPathFile.getName();

        return atomicRenameEntry(oldParent, oldName, newParent, newName);

@@ -135,9 +128,9 @@ public class FilerClient {

    public boolean rm(String path, boolean isRecursive, boolean ignoreRecusiveError) {

-        Path pathObject = Paths.get(path);
-        String parent = pathObject.getParent().toString();
-        String name = pathObject.getFileName().toString();
+        File pathFile = new File(path);
+        String parent = pathFile.getParent();
+        String name = pathFile.getName();

        return deleteEntry(
                parent,
@@ -154,9 +147,9 @@ public class FilerClient {

    public boolean touch(String path, int mode, int uid, int gid, String userName, String[] groupNames) {

-        Path pathObject = Paths.get(path);
-        String parent = pathObject.getParent().toString();
-        String name = pathObject.getFileName().toString();
+        File pathFile = new File(path);
+        String parent = pathFile.getParent();
+        String name = pathFile.getName();

        FilerProto.Entry entry = lookupEntry(parent, name);
        if (entry == null) {
@@ -236,7 +229,7 @@ public class FilerClient {
    }

    public List<FilerProto.Entry> listEntries(String path, String entryPrefix, String lastEntryName, int limit, boolean includeLastEntry) {
-        Iterator<FilerProto.ListEntriesResponse> iter = filerGrpcClient.getBlockingStub().listEntries(FilerProto.ListEntriesRequest.newBuilder()
+        Iterator<FilerProto.ListEntriesResponse> iter = this.getBlockingStub().listEntries(FilerProto.ListEntriesRequest.newBuilder()
                .setDirectory(path)
                .setPrefix(entryPrefix)
                .setStartFromFileName(lastEntryName)
@@ -253,7 +246,7 @@ public class FilerClient {

    public FilerProto.Entry lookupEntry(String directory, String entryName) {
        try {
-            FilerProto.Entry entry = filerGrpcClient.getBlockingStub().lookupDirectoryEntry(
+            FilerProto.Entry entry = this.getBlockingStub().lookupDirectoryEntry(
                    FilerProto.LookupDirectoryEntryRequest.newBuilder()
                            .setDirectory(directory)
                            .setName(entryName)
@@ -274,10 +267,10 @@ public class FilerClient {
    public boolean createEntry(String parent, FilerProto.Entry entry) {
        try {
            FilerProto.CreateEntryResponse createEntryResponse =
-                    filerGrpcClient.getBlockingStub().createEntry(FilerProto.CreateEntryRequest.newBuilder()
-                    .setDirectory(parent)
-                    .setEntry(entry)
-                    .build());
+                    this.getBlockingStub().createEntry(FilerProto.CreateEntryRequest.newBuilder()
+                            .setDirectory(parent)
+                            .setEntry(entry)
+                            .build());
            if (Strings.isNullOrEmpty(createEntryResponse.getError())) {
                return true;
            }
@@ -291,7 +284,7 @@ public class FilerClient {

    public boolean updateEntry(String parent, FilerProto.Entry entry) {
        try {
-            filerGrpcClient.getBlockingStub().updateEntry(FilerProto.UpdateEntryRequest.newBuilder()
+            this.getBlockingStub().updateEntry(FilerProto.UpdateEntryRequest.newBuilder()
                    .setDirectory(parent)
                    .setEntry(entry)
                    .build());
@@ -304,7 +297,7 @@ public class FilerClient {

    public boolean deleteEntry(String parent, String entryName, boolean isDeleteFileChunk, boolean isRecursive, boolean ignoreRecusiveError) {
        try {
-            filerGrpcClient.getBlockingStub().deleteEntry(FilerProto.DeleteEntryRequest.newBuilder()
+            this.getBlockingStub().deleteEntry(FilerProto.DeleteEntryRequest.newBuilder()
                    .setDirectory(parent)
                    .setName(entryName)
                    .setIsDeleteData(isDeleteFileChunk)
@@ -320,7 +313,7 @@ public class FilerClient {

    public boolean atomicRenameEntry(String oldParent, String oldName, String newParent, String newName) {
        try {
-            filerGrpcClient.getBlockingStub().atomicRenameEntry(FilerProto.AtomicRenameEntryRequest.newBuilder()
+            this.getBlockingStub().atomicRenameEntry(FilerProto.AtomicRenameEntryRequest.newBuilder()
                    .setOldDirectory(oldParent)
                    .setOldName(oldName)
                    .setNewDirectory(newParent)
@@ -333,4 +326,13 @@ public class FilerClient {
        return true;
    }

+    public Iterator<FilerProto.SubscribeMetadataResponse> watch(String prefix, String clientName, long sinceNs) {
+        return this.getBlockingStub().subscribeMetadata(FilerProto.SubscribeMetadataRequest.newBuilder()
+                .setPathPrefix(prefix)
+                .setClientName(clientName)
+                .setSinceNs(sinceNs)
+                .build()
+        );
+    }
+
 }
@@ -9,8 +9,8 @@ import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;

 import javax.net.ssl.SSLException;
-import java.util.Map;
 import java.util.HashMap;
+import java.util.Map;
 import java.util.concurrent.TimeUnit;

 public class FilerGrpcClient {
@@ -26,6 +26,9 @@ public class FilerGrpcClient {
        }
    }

+    public final int VOLUME_SERVER_ACCESS_DIRECT = 0;
+    public final int VOLUME_SERVER_ACCESS_PUBLIC_URL = 1;
+    public final int VOLUME_SERVER_ACCESS_FILER_PROXY = 2;
    public final Map<String, FilerProto.Locations> vidLocations = new HashMap<>();
    private final ManagedChannel channel;
    private final SeaweedFilerGrpc.SeaweedFilerBlockingStub blockingStub;
@@ -34,6 +37,8 @@ public class FilerGrpcClient {
    private boolean cipher = false;
    private String collection = "";
    private String replication = "";
+    private int volumeServerAccess = VOLUME_SERVER_ACCESS_DIRECT;
+    private String filerAddress;

    public FilerGrpcClient(String host, int grpcPort) {
        this(host, grpcPort, sslContext);
@@ -49,6 +54,8 @@ public class FilerGrpcClient {
                        .negotiationType(NegotiationType.TLS)
                        .sslContext(sslContext));

+        filerAddress = String.format("%s:%d", host, grpcPort - 10000);
+
        FilerProto.GetFilerConfigurationResponse filerConfigurationResponse =
                this.getBlockingStub().getFilerConfiguration(
                        FilerProto.GetFilerConfigurationRequest.newBuilder().build());
@@ -58,7 +65,7 @@ public class FilerGrpcClient {

    }

-    public FilerGrpcClient(ManagedChannelBuilder<?> channelBuilder) {
+    private FilerGrpcClient(ManagedChannelBuilder<?> channelBuilder) {
        channel = channelBuilder.build();
        blockingStub = SeaweedFilerGrpc.newBlockingStub(channel);
        asyncStub = SeaweedFilerGrpc.newStub(channel);
@@ -93,4 +100,39 @@ public class FilerGrpcClient {
        return futureStub;
    }

+    public void setAccessVolumeServerDirectly() {
+        this.volumeServerAccess = VOLUME_SERVER_ACCESS_DIRECT;
+    }
+
+    public boolean isAccessVolumeServerDirectly() {
+        return this.volumeServerAccess == VOLUME_SERVER_ACCESS_DIRECT;
+    }
+
+    public void setAccessVolumeServerByPublicUrl() {
+        this.volumeServerAccess = VOLUME_SERVER_ACCESS_PUBLIC_URL;
+    }
+
+    public boolean isAccessVolumeServerByPublicUrl() {
+        return this.volumeServerAccess == VOLUME_SERVER_ACCESS_PUBLIC_URL;
+    }
+
+    public void setAccessVolumeServerByFilerProxy() {
+        this.volumeServerAccess = VOLUME_SERVER_ACCESS_FILER_PROXY;
+    }
+
+    public boolean isAccessVolumeServerByFilerProxy() {
+        return this.volumeServerAccess == VOLUME_SERVER_ACCESS_FILER_PROXY;
+    }
+
+    public String getChunkUrl(String chunkId, String url, String publicUrl) {
+        switch (this.volumeServerAccess) {
+            case VOLUME_SERVER_ACCESS_PUBLIC_URL:
+                return String.format("http://%s/%s", publicUrl, chunkId);
+            case VOLUME_SERVER_ACCESS_FILER_PROXY:
+                return String.format("http://%s/?proxyChunkId=%s", this.filerAddress, chunkId);
+            default:
+                return String.format("http://%s/%s", url, chunkId);
+        }
+    }
+
 }
@@ -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 FilerClient filerClient;
    private final String path;
    private final FilerProto.Entry entry;
    private final List<SeaweedRead.VisibleInterval> visibleIntervalList;
@@ -31,17 +27,31 @@ public class SeaweedInputStream extends FSInputStream {
    private boolean closed = false;

    public SeaweedInputStream(
-            final FilerGrpcClient filerGrpcClient,
-            final Statistics statistics,
+            final FilerClient filerClient,
+            final String fullpath) throws IOException {
+        this.path = fullpath;
+        this.filerClient = filerClient;
+        this.entry = filerClient.lookupEntry(
+                SeaweedOutputStream.getParentDirectory(fullpath),
+                SeaweedOutputStream.getFileName(fullpath));
+        this.contentLength = SeaweedRead.fileSize(entry);
+
+        this.visibleIntervalList = SeaweedRead.nonOverlappingVisibleIntervals(filerClient, entry.getChunksList());
+
+        LOG.debug("new path:{} entry:{} visibleIntervalList:{}", path, entry, visibleIntervalList);
+
+    }
+
+    public SeaweedInputStream(
+            final FilerClient filerClient,
            final String path,
            final FilerProto.Entry entry) throws IOException {
-        this.filerGrpcClient = filerGrpcClient;
-        this.statistics = statistics;
+        this.filerClient = filerClient;
        this.path = path;
        this.entry = entry;
        this.contentLength = SeaweedRead.fileSize(entry);

-        this.visibleIntervalList = SeaweedRead.nonOverlappingVisibleIntervals(filerGrpcClient, entry.getChunksList());
+        this.visibleIntervalList = SeaweedRead.nonOverlappingVisibleIntervals(filerClient, entry.getChunksList());

        LOG.debug("new path:{} entry:{} visibleIntervalList:{}", path, entry, visibleIntervalList);

@@ -63,14 +73,8 @@ 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 +88,62 @@ 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.filerClient, 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 +171,9 @@ 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 +188,21 @@ 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;
@@ -18,21 +11,19 @@ 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;
+    protected final boolean supportFlush = true;
+    private final FilerClient filerClient;
+    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;
    private final ConcurrentLinkedDeque<WriteOperation> writeOperations;
+    private final boolean shouldSaveMetadata = false;
+    private FilerProto.Entry.Builder entry;
    private long position;
    private boolean closed;
    private volatile IOException lastError;
@@ -42,9 +33,17 @@ 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(FilerClient filerClient, final String fullpath) {
+        this(filerClient, fullpath, "000");
+    }
+
+    public SeaweedOutputStream(FilerClient filerClient, final String fullpath, final String replication) {
+        this(filerClient, fullpath, null, 0, 8 * 1024 * 1024, "000");
+    }
+
+    public SeaweedOutputStream(FilerClient filerClient, final String path, FilerProto.Entry.Builder entry,
                               final long position, final int bufferSize, final String replication) {
-        this.filerGrpcClient = filerGrpcClient;
+        this.filerClient = filerClient;
        this.replication = replication;
        this.path = path;
        this.position = position;
@@ -67,12 +66,50 @@ public class SeaweedOutputStream extends OutputStream {
        this.completionService = new ExecutorCompletionService<>(this.threadExecutor);

        this.entry = entry;
+        if (this.entry == null) {
+            long now = System.currentTimeMillis() / 1000L;

+            this.entry = FilerProto.Entry.newBuilder()
+                    .setName(getFileName(path))
+                    .setIsDirectory(false)
+                    .setAttributes(FilerProto.FuseAttributes.newBuilder()
+                            .setFileMode(0755)
+                            .setReplication(replication)
+                            .setCrtime(now)
+                            .setMtime(now)
+                            .clearGroupName()
+                    );
+        }
+
+    }
+
+    public static String getParentDirectory(String path) {
+        int protoIndex = path.indexOf("://");
+        if (protoIndex >= 0) {
+            int pathStart = path.indexOf("/", protoIndex+3);
+            path = path.substring(pathStart);
+        }
+        if (path.equals("/")) {
+            return path;
+        }
+        int lastSlashIndex = path.lastIndexOf("/");
+        if (lastSlashIndex == 0) {
+            return "/";
+        }
+        return path.substring(0, lastSlashIndex);
+    }
+
+    public static String getFileName(String path) {
+        if (path.indexOf("/") < 0) {
+            return path;
+        }
+        int lastSlashIndex = path.lastIndexOf("/");
+        return path.substring(lastSlashIndex + 1);
    }

    private synchronized void flushWrittenBytesToServiceInternal(final long offset) throws IOException {
        try {
-            SeaweedWrite.writeMeta(filerGrpcClient, getParentDirectory(path), entry);
+            SeaweedWrite.writeMeta(filerClient, getParentDirectory(path), entry);
        } catch (Exception ex) {
            throw new IOException(ex);
        }
@@ -89,7 +126,9 @@ 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 +191,7 @@ 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;
@@ -162,6 +201,7 @@ public class SeaweedOutputStream extends OutputStream {
                threadExecutor.shutdownNow();
            }
        }
+
    }

    private synchronized void writeCurrentBufferToService() throws IOException {
@@ -185,7 +225,7 @@ 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());
+            SeaweedWrite.writeData(entry, replication, filerClient, writePosition, bufferToWrite.array(), bufferToWrite.position(), bufferToWrite.limit(), path);
            // System.out.println(path + " saved [" + (writePosition) + "," + ((writePosition) + bytesLength) + ")");
            ByteBufferPool.release(bufferToWrite);
            return null;
@@ -239,13 +279,13 @@ 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 +318,6 @@ 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 {
@@ -19,24 +20,36 @@ public class SeaweedRead {
    private static final Logger LOG = LoggerFactory.getLogger(SeaweedRead.class);

    static ChunkCache chunkCache = new ChunkCache(4);
+    static VolumeIdCache volumeIdCache = new VolumeIdCache(4 * 1024);

    // 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 {
+    public static long read(FilerClient filerClient, List<VisibleInterval> visibleIntervals,
+                            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<>();

        FilerProto.LookupVolumeRequest.Builder lookupRequest = FilerProto.LookupVolumeRequest.newBuilder();
        for (ChunkView chunkView : chunkViews) {
            String vid = parseVolumeId(chunkView.fileId);
-            lookupRequest.addVolumeIds(vid);
+            FilerProto.Locations locations = volumeIdCache.getLocations(vid);
+            if (locations == null) {
+                lookupRequest.addVolumeIds(vid);
+            } else {
+                knownLocations.put(vid, locations);
+            }
        }

-        FilerProto.LookupVolumeResponse lookupResponse = filerGrpcClient
-                .getBlockingStub().lookupVolume(lookupRequest.build());
-
-        Map<String, FilerProto.Locations> vid2Locations = lookupResponse.getLocationsMapMap();
+        if (lookupRequest.getVolumeIdsCount() > 0) {
+            FilerProto.LookupVolumeResponse lookupResponse = filerClient
+                    .getBlockingStub().lookupVolume(lookupRequest.build());
+            Map<String, FilerProto.Locations> vid2Locations = lookupResponse.getLocationsMapMap();
+            for (Map.Entry<String, FilerProto.Locations> entry : vid2Locations.entrySet()) {
+                volumeIdCache.setLocations(entry.getKey(), entry.getValue());
+                knownLocations.put(entry.getKey(), entry.getValue());
+            }
+        }

        //TODO parallel this
        long readCount = 0;
@@ -46,18 +59,19 @@ 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;
            }

-            FilerProto.Locations locations = vid2Locations.get(parseVolumeId(chunkView.fileId));
+            FilerProto.Locations locations = knownLocations.get(parseVolumeId(chunkView.fileId));
            if (locations == null || locations.getLocationsCount() == 0) {
                LOG.error("failed to locate {}", chunkView.fileId);
                // log here!
                return 0;
            }

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

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

@@ -66,11 +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;
        }
@@ -78,30 +93,30 @@ 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(FilerClient filerClient, long startOffset, ByteBuffer buf, ChunkView chunkView, FilerProto.Locations locations) throws IOException {

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

        if (chunkData == null) {
-            chunkData = doFetchFullChunkData(chunkView, locations);
+            chunkData = doFetchFullChunkData(filerClient, chunkView, locations);
            chunkCache.setChunk(chunkView.fileId, chunkData);
        }

        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;
    }

-    public static byte[] doFetchFullChunkData(ChunkView chunkView, FilerProto.Locations locations) throws IOException {
+    public static byte[] doFetchFullChunkData(FilerClient filerClient, ChunkView chunkView, FilerProto.Locations locations) throws IOException {

        byte[] data = null;
        IOException lastException = null;
        for (long waitTime = 1000L; waitTime < 10 * 1000; waitTime += waitTime / 2) {
            for (FilerProto.Location location : locations.getLocationsList()) {
-                String url = String.format("http://%s/%s", location.getUrl(), chunkView.fileId);
+                String url = filerClient.getChunkUrl(chunkView.fileId, location.getUrl(), location.getPublicUrl());
                try {
                    data = doFetchOneFullChunkData(chunkView, url);
                    lastException = null;
@@ -130,7 +145,7 @@ public class SeaweedRead {

    }

-    public static byte[] doFetchOneFullChunkData(ChunkView chunkView, String url) throws IOException {
+    private static byte[] doFetchOneFullChunkData(ChunkView chunkView, String url) throws IOException {

        HttpGet request = new HttpGet(url);

@@ -206,9 +221,9 @@ public class SeaweedRead {
    }

    public static List<VisibleInterval> nonOverlappingVisibleIntervals(
-            final FilerGrpcClient filerGrpcClient, List<FilerProto.FileChunk> chunkList) throws IOException {
+            final FilerClient filerClient, List<FilerProto.FileChunk> chunkList) throws IOException {

-        chunkList = FileChunkManifest.resolveChunkManifest(filerGrpcClient, chunkList);
+        chunkList = FileChunkManifest.resolveChunkManifest(filerClient, chunkList);

        FilerProto.FileChunk[] chunks = chunkList.toArray(new FilerProto.FileChunk[0]);
        Arrays.sort(chunks, new Comparator<FilerProto.FileChunk>() {
@@ -23,38 +23,41 @@ public class SeaweedWrite {

    public static void writeData(FilerProto.Entry.Builder entry,
                                 final String replication,
-                                 final FilerGrpcClient filerGrpcClient,
+                                 final FilerClient filerClient,
                                 final long offset,
                                 final byte[] bytes,
-                                 final long bytesOffset, final long bytesLength) throws IOException {
+                                 final long bytesOffset, final long bytesLength,
+                                 final String path) throws IOException {
        FilerProto.FileChunk.Builder chunkBuilder = writeChunk(
-                replication, filerGrpcClient, offset, bytes, bytesOffset, bytesLength);
+                replication, filerClient, offset, bytes, bytesOffset, bytesLength, path);
        synchronized (entry) {
            entry.addChunks(chunkBuilder);
        }
    }

    public static FilerProto.FileChunk.Builder writeChunk(final String replication,
-                                                          final FilerGrpcClient filerGrpcClient,
+                                                          final FilerClient filerClient,
                                                          final long offset,
                                                          final byte[] bytes,
                                                          final long bytesOffset,
-                                                          final long bytesLength) throws IOException {
-        FilerProto.AssignVolumeResponse response = filerGrpcClient.getBlockingStub().assignVolume(
+                                                          final long bytesLength,
+                                                          final String path) throws IOException {
+        FilerProto.AssignVolumeResponse response = filerClient.getBlockingStub().assignVolume(
                FilerProto.AssignVolumeRequest.newBuilder()
-                        .setCollection(filerGrpcClient.getCollection())
-                        .setReplication(replication == null ? filerGrpcClient.getReplication() : replication)
+                        .setCollection(filerClient.getCollection())
+                        .setReplication(replication == null ? filerClient.getReplication() : replication)
                        .setDataCenter("")
                        .setTtlSec(0)
+                        .setPath(path)
                        .build());
        String fileId = response.getFileId();
-        String url = response.getUrl();
        String auth = response.getAuth();
-        String targetUrl = String.format("http://%s/%s", url, fileId);
+
+        String targetUrl = filerClient.getChunkUrl(fileId, response.getUrl(), response.getPublicUrl());

        ByteString cipherKeyString = com.google.protobuf.ByteString.EMPTY;
        byte[] cipherKey = null;
-        if (filerGrpcClient.isCipher()) {
+        if (filerClient.isCipher()) {
            cipherKey = genCipherKey();
            cipherKeyString = ByteString.copyFrom(cipherKey);
        }
@@ -72,15 +75,15 @@ public class SeaweedWrite {
                .setCipherKey(cipherKeyString);
    }

-    public static void writeMeta(final FilerGrpcClient filerGrpcClient,
+    public static void writeMeta(final FilerClient filerClient,
                                 final String parentDirectory,
                                 final FilerProto.Entry.Builder entry) throws IOException {

        synchronized (entry) {
-            List<FilerProto.FileChunk> chunks = FileChunkManifest.maybeManifestize(filerGrpcClient, entry.getChunksList());
+            List<FilerProto.FileChunk> chunks = FileChunkManifest.maybeManifestize(filerClient, entry.getChunksList(), parentDirectory);
            entry.clearChunks();
            entry.addAllChunks(chunks);
-            filerGrpcClient.getBlockingStub().createEntry(
+            filerClient.getBlockingStub().createEntry(
                    FilerProto.CreateEntryRequest.newBuilder()
                            .setDirectory(parentDirectory)
                            .setEntry(entry)
@@ -0,0 +1,36 @@
+package seaweedfs.client;
+
+import com.google.common.cache.Cache;
+import com.google.common.cache.CacheBuilder;
+
+import java.util.concurrent.TimeUnit;
+
+public class VolumeIdCache {
+
+    private Cache<String, FilerProto.Locations> cache = null;
+
+    public VolumeIdCache(int maxEntries) {
+        if (maxEntries == 0) {
+            return;
+        }
+        this.cache = CacheBuilder.newBuilder()
+                .maximumSize(maxEntries)
+                .expireAfterAccess(5, TimeUnit.MINUTES)
+                .build();
+    }
+
+    public FilerProto.Locations getLocations(String volumeId) {
+        if (this.cache == null) {
+            return null;
+        }
+        return this.cache.getIfPresent(volumeId);
+    }
+
+    public void setLocations(String volumeId, FilerProto.Locations locations) {
+        if (this.cache == null) {
+            return;
+        }
+        this.cache.put(volumeId, locations);
+    }
+
+}
@@ -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 {
@@ -155,6 +156,7 @@ message FuseAttributes {
    repeated string group_name = 12; // for hdfs
    string symlink_target = 13;
    bytes md5 = 14;
+    string disk_type = 15;
 }

 message CreateEntryRequest {
@@ -217,8 +219,9 @@ message AssignVolumeRequest {
    string replication = 3;
    int32 ttl_sec = 4;
    string data_center = 5;
-    string parent_path = 6;
+    string path = 6;
    string rack = 7;
+    string disk_type = 8;
 }

 message AssignVolumeResponse {
@@ -269,11 +272,9 @@ message StatisticsRequest {
    string replication = 1;
    string collection = 2;
    string ttl = 3;
+    string disk_type = 4;
 }
 message StatisticsResponse {
-    string replication = 1;
-    string collection = 2;
-    string ttl = 3;
    uint64 total_size = 4;
    uint64 used_size = 5;
    uint64 file_count = 6;
@@ -348,3 +349,18 @@ message KvPutRequest {
 message KvPutResponse {
    string error = 1;
 }
+
+// path-based configurations
+message FilerConf {
+    int32 version = 1;
+    message PathConf {
+        string location_prefix = 1;
+        string collection = 2;
+        string replication = 3;
+        string ttl = 4;
+        string disk_type = 5;
+        bool fsync = 6;
+        uint32 volume_growth_count = 7;
+    }
+    repeated PathConf locations = 2;
+}
@@ -0,0 +1,32 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<project xmlns="http://maven.apache.org/POM/4.0.0"
+         xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+         xsi:schemaLocation="http://maven.apache.org/POM/4.0.0 http://maven.apache.org/xsd/maven-4.0.0.xsd">
+    <modelVersion>4.0.0</modelVersion>
+
+    <groupId>org.example</groupId>
+    <artifactId>unzip</artifactId>
+    <version>1.0-SNAPSHOT</version>
+    <dependencies>
+        <dependency>
+            <groupId>com.github.chrislusf</groupId>
+            <artifactId>seaweedfs-client</artifactId>
+            <version>1.6.1</version>
+            <scope>compile</scope>
+        </dependency>
+        <dependency>
+            <groupId>com.github.chrislusf</groupId>
+            <artifactId>seaweedfs-hadoop2-client</artifactId>
+            <version>1.6.1</version>
+            <scope>compile</scope>
+        </dependency>
+        <dependency>
+            <groupId>org.apache.hadoop</groupId>
+            <artifactId>hadoop-common</artifactId>
+            <version>2.9.2</version>
+            <scope>compile</scope>
+        </dependency>
+    </dependencies>
+
+
+</project>
@@ -0,0 +1,48 @@
+package com.seaweedfs.examples;
+
+import seaweedfs.client.FilerClient;
+import seaweedfs.client.SeaweedInputStream;
+
+import java.io.FileInputStream;
+import java.io.IOException;
+import java.io.InputStream;
+import java.util.zip.ZipEntry;
+import java.util.zip.ZipInputStream;
+
+public class ExampleReadFile {
+
+    public static void main(String[] args) throws IOException {
+
+        FilerClient filerClient = new FilerClient("localhost", 18888);
+
+        long startTime = System.currentTimeMillis();
+        parseZip("/Users/chris/tmp/test.zip");
+
+        long startTime2 = System.currentTimeMillis();
+
+        long localProcessTime = startTime2 - startTime;
+
+        SeaweedInputStream seaweedInputStream = new SeaweedInputStream(
+                filerClient, "/test.zip");
+        parseZip(seaweedInputStream);
+
+        long swProcessTime = System.currentTimeMillis() - startTime2;
+
+        System.out.println("Local time: " + localProcessTime);
+        System.out.println("SeaweedFS time: " + swProcessTime);
+
+    }
+
+    public static void parseZip(String filename) throws IOException {
+        FileInputStream fileInputStream = new FileInputStream(filename);
+        parseZip(fileInputStream);
+    }
+
+    public static void parseZip(InputStream is) throws IOException {
+        ZipInputStream zin = new ZipInputStream(is);
+        ZipEntry ze;
+        while ((ze = zin.getNextEntry()) != null) {
+            System.out.println(ze.getName());
+        }
+    }
+}
@@ -0,0 +1,46 @@
+package com.seaweedfs.examples;
+
+import seaweedfs.client.FilerClient;
+import seaweedfs.client.FilerProto;
+
+import java.io.IOException;
+import java.util.Date;
+import java.util.Iterator;
+
+public class ExampleWatchFileChanges {
+
+    public static void main(String[] args) throws IOException {
+
+        FilerClient filerClient = new FilerClient("localhost", 18888);
+
+        long sinceNs = (System.currentTimeMillis() - 3600 * 1000) * 1000000L;
+
+        Iterator<FilerProto.SubscribeMetadataResponse> watch = filerClient.watch(
+                "/buckets",
+                "exampleClientName",
+                sinceNs
+        );
+
+        System.out.println("Connected to filer, subscribing from " + new Date());
+
+        while (watch.hasNext()) {
+            FilerProto.SubscribeMetadataResponse event = watch.next();
+            FilerProto.EventNotification notification = event.getEventNotification();
+            if (!event.getDirectory().equals(notification.getNewParentPath())) {
+                // move an entry to a new directory, possibly with a new name
+                if (notification.hasOldEntry() && notification.hasNewEntry()) {
+                    System.out.println("moved " + event.getDirectory() + "/" + notification.getOldEntry().getName() + " to " + notification.getNewParentPath() + "/" + notification.getNewEntry().getName());
+                } else {
+                    System.out.println("this should not happen.");
+                }
+            } else if (notification.hasNewEntry() && !notification.hasOldEntry()) {
+                System.out.println("created entry " + event.getDirectory() + "/" + notification.getNewEntry().getName());
+            } else if (!notification.hasNewEntry() && notification.hasOldEntry()) {
+                System.out.println("deleted entry " + event.getDirectory() + "/" + notification.getOldEntry().getName());
+            } else if (notification.hasNewEntry() && notification.hasOldEntry()) {
+                System.out.println("updated entry " + event.getDirectory() + "/" + notification.getNewEntry().getName());
+            }
+        }
+
+    }
+}
@@ -0,0 +1,47 @@
+package com.seaweedfs.examples;
+
+import seaweedfs.client.FilerClient;
+import seaweedfs.client.SeaweedInputStream;
+import seaweedfs.client.SeaweedOutputStream;
+
+import java.io.IOException;
+import java.io.InputStream;
+import java.util.zip.ZipEntry;
+import java.util.zip.ZipInputStream;
+
+public class ExampleWriteFile {
+
+    public static void main(String[] args) throws IOException {
+
+        FilerClient filerClient = new FilerClient("localhost", 18888);
+
+        SeaweedInputStream seaweedInputStream = new SeaweedInputStream(filerClient, "/test.zip");
+        unZipFiles(filerClient, seaweedInputStream);
+
+    }
+
+    public static void unZipFiles(FilerClient filerClient, InputStream is) throws IOException {
+        ZipInputStream zin = new ZipInputStream(is);
+        ZipEntry ze;
+        while ((ze = zin.getNextEntry()) != null) {
+
+            String filename = ze.getName();
+            if (filename.indexOf("/") >= 0) {
+                filename = filename.substring(filename.lastIndexOf("/") + 1);
+            }
+            if (filename.length()==0) {
+                continue;
+            }
+
+            SeaweedOutputStream seaweedOutputStream = new SeaweedOutputStream(filerClient, "/test/"+filename);
+            byte[] bytesIn = new byte[16 * 1024];
+            int read = 0;
+            while ((read = zin.read(bytesIn))!=-1) {
+                seaweedOutputStream.write(bytesIn,0,read);
+            }
+            seaweedOutputStream.close();
+
+            System.out.println(ze.getName());
+        }
+    }
+}
@@ -0,0 +1,25 @@
+package com.seaweedfs.examples;
+
+import org.apache.hadoop.conf.Configuration;
+import org.apache.hadoop.fs.FileSystem;
+import org.apache.hadoop.fs.Path;
+import org.apache.hadoop.io.IOUtils;
+
+import java.io.*;
+
+public class HdfsCopyFile {
+    public static void main(String[] args) throws IOException {
+        Configuration configuration = new Configuration();
+
+        configuration.set("fs.defaultFS", "seaweedfs://localhost:8888");
+        configuration.set("fs.seaweedfs.impl", "seaweed.hdfs.SeaweedFileSystem");
+
+        FileSystem fs = FileSystem.get(configuration);
+        String source = "/Users/chris/tmp/test.zip";
+        String destination = "/buckets/spark/test01.zip";
+        InputStream in = new BufferedInputStream(new FileInputStream(source));
+
+        OutputStream out = fs.create(new Path(destination));
+        IOUtils.copyBytes(in, out, 4096, true);
+    }
+}
@@ -301,7 +301,7 @@
    </snapshotRepository>
  </distributionManagement>
  <properties>
-    <seaweedfs.client.version>1.5.1</seaweedfs.client.version>
+    <seaweedfs.client.version>1.6.1</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>1.5.1</seaweedfs.client.version>
+        <seaweedfs.client.version>1.6.1</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");
+        }
+    }
+}
@@ -25,6 +25,8 @@ public class SeaweedFileSystem extends FileSystem {
    public static final String FS_SEAWEED_FILER_PORT = "fs.seaweed.filer.port";
    public static final int FS_SEAWEED_DEFAULT_PORT = 8888;
    public static final String FS_SEAWEED_BUFFER_SIZE = "fs.seaweed.buffer.size";
+    public static final String FS_SEAWEED_REPLICATION = "fs.seaweed.replication";
+    public static final String FS_SEAWEED_VOLUME_SERVER_ACCESS = "fs.seaweed.volume.server.access";
    public static final int FS_SEAWEED_DEFAULT_BUFFER_SIZE = 4 * 1024 * 1024;

    private static final Logger LOG = LoggerFactory.getLogger(SeaweedFileSystem.class);
@@ -65,6 +67,12 @@ public class SeaweedFileSystem extends FileSystem {

    }

+    @Override
+    public void close() throws IOException {
+        super.close();
+        this.seaweedFileSystemStore.close();
+    }
+
    @Override
    public FSDataInputStream open(Path path, int bufferSize) throws IOException {

@@ -75,7 +83,7 @@ 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;
@@ -91,7 +99,7 @@ public class SeaweedFileSystem extends FileSystem {
        path = qualify(path);

        try {
-            String replicaPlacement = String.format("%03d", replication - 1);
+            String replicaPlacement = this.getConf().get(FS_SEAWEED_REPLICATION, String.format("%03d", replication - 1));
            int seaweedBufferSize = this.getConf().getInt(FS_SEAWEED_BUFFER_SIZE, FS_SEAWEED_DEFAULT_BUFFER_SIZE);
            OutputStream outputStream = seaweedFileSystemStore.createFile(path, overwrite, permission, seaweedBufferSize, replicaPlacement);
            return new FSDataOutputStream(outputStream, statistics);
@@ -18,22 +18,34 @@ import java.util.ArrayList;
 import java.util.Arrays;
 import java.util.List;

-import static seaweed.hdfs.SeaweedFileSystem.FS_SEAWEED_BUFFER_SIZE;
-import static seaweed.hdfs.SeaweedFileSystem.FS_SEAWEED_DEFAULT_BUFFER_SIZE;
+import static seaweed.hdfs.SeaweedFileSystem.*;

 public class SeaweedFileSystemStore {

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

-    private FilerGrpcClient filerGrpcClient;
    private FilerClient filerClient;
    private Configuration conf;

    public SeaweedFileSystemStore(String host, int port, Configuration conf) {
        int grpcPort = 10000 + port;
-        filerGrpcClient = new FilerGrpcClient(host, grpcPort);
-        filerClient = new FilerClient(filerGrpcClient);
+        filerClient = new FilerClient(host, grpcPort);
        this.conf = conf;
+        String volumeServerAccessMode = this.conf.get(FS_SEAWEED_VOLUME_SERVER_ACCESS, "direct");
+        if (volumeServerAccessMode.equals("publicUrl")) {
+            filerClient.setAccessVolumeServerByPublicUrl();
+        } else if (volumeServerAccessMode.equals("filerProxy")) {
+            filerClient.setAccessVolumeServerByFilerProxy();
+        }
+
+    }
+
+    public void close() {
+        try {
+            this.filerClient.shutdown();
+        } catch (InterruptedException e) {
+            e.printStackTrace();
+        }
    }

    public static String getParentDirectory(Path path) {
@@ -185,6 +197,7 @@ 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);
@@ -204,10 +217,10 @@ public class SeaweedFileSystemStore {
                    .clearGroupName()
                    .addAllGroupName(Arrays.asList(userGroupInformation.getGroupNames()))
                );
-            SeaweedWrite.writeMeta(filerGrpcClient, getParentDirectory(path), entry);
+            SeaweedWrite.writeMeta(filerClient, getParentDirectory(path), entry);
        }

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

    }

@@ -221,7 +234,7 @@ public class SeaweedFileSystemStore {
            throw new FileNotFoundException("read non-exist file " + path);
        }

-        return new SeaweedInputStream(filerGrpcClient,
+        return new SeaweedHadoopInputStream(filerClient,
            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.FilerClient;
+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 FilerClient filerClient,
+            final Statistics statistics,
+            final String path,
+            final FilerProto.Entry entry) throws IOException {
+        this.seaweedInputStream = new SeaweedInputStream(filerClient, 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.FilerClient;
+import seaweedfs.client.FilerProto;
+import seaweedfs.client.SeaweedOutputStream;
+
+public class SeaweedHadoopOutputStream extends SeaweedOutputStream {
+
+    public SeaweedHadoopOutputStream(FilerClient filerClient, final String path, FilerProto.Entry.Builder entry,
+                                     final long position, final int bufferSize, final String replication) {
+        super(filerClient, path, entry, position, bufferSize, replication);
+    }
+
+}
@@ -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;
-    }
-}
@@ -309,7 +309,7 @@
    </snapshotRepository>
  </distributionManagement>
  <properties>
-    <seaweedfs.client.version>1.5.1</seaweedfs.client.version>
+    <seaweedfs.client.version>1.6.1</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>1.5.1</seaweedfs.client.version>
+        <seaweedfs.client.version>1.6.1</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");
+        }
+    }
+}
@@ -25,6 +25,8 @@ public class SeaweedFileSystem extends FileSystem {
    public static final String FS_SEAWEED_FILER_PORT = "fs.seaweed.filer.port";
    public static final int FS_SEAWEED_DEFAULT_PORT = 8888;
    public static final String FS_SEAWEED_BUFFER_SIZE = "fs.seaweed.buffer.size";
+    public static final String FS_SEAWEED_REPLICATION = "fs.seaweed.replication";
+    public static final String FS_SEAWEED_VOLUME_SERVER_ACCESS = "fs.seaweed.volume.server.access";
    public static final int FS_SEAWEED_DEFAULT_BUFFER_SIZE = 4 * 1024 * 1024;

    private static final Logger LOG = LoggerFactory.getLogger(SeaweedFileSystem.class);
@@ -65,6 +67,12 @@ public class SeaweedFileSystem extends FileSystem {

    }

+    @Override
+    public void close() throws IOException {
+        super.close();
+        this.seaweedFileSystemStore.close();
+    }
+
    @Override
    public FSDataInputStream open(Path path, int bufferSize) throws IOException {

@@ -75,7 +83,7 @@ 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;
@@ -91,7 +99,7 @@ public class SeaweedFileSystem extends FileSystem {
        path = qualify(path);

        try {
-            String replicaPlacement = String.format("%03d", replication - 1);
+            String replicaPlacement = this.getConf().get(FS_SEAWEED_REPLICATION, String.format("%03d", replication - 1));
            int seaweedBufferSize = this.getConf().getInt(FS_SEAWEED_BUFFER_SIZE, FS_SEAWEED_DEFAULT_BUFFER_SIZE);
            OutputStream outputStream = seaweedFileSystemStore.createFile(path, overwrite, permission, seaweedBufferSize, replicaPlacement);
            return new FSDataOutputStream(outputStream, statistics);
@@ -18,22 +18,34 @@ import java.util.ArrayList;
 import java.util.Arrays;
 import java.util.List;

-import static seaweed.hdfs.SeaweedFileSystem.FS_SEAWEED_BUFFER_SIZE;
-import static seaweed.hdfs.SeaweedFileSystem.FS_SEAWEED_DEFAULT_BUFFER_SIZE;
+import static seaweed.hdfs.SeaweedFileSystem.*;

 public class SeaweedFileSystemStore {

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

-    private FilerGrpcClient filerGrpcClient;
    private FilerClient filerClient;
    private Configuration conf;

    public SeaweedFileSystemStore(String host, int port, Configuration conf) {
        int grpcPort = 10000 + port;
-        filerGrpcClient = new FilerGrpcClient(host, grpcPort);
-        filerClient = new FilerClient(filerGrpcClient);
+        filerClient = new FilerClient(host, grpcPort);
        this.conf = conf;
+        String volumeServerAccessMode = this.conf.get(FS_SEAWEED_VOLUME_SERVER_ACCESS, "direct");
+        if (volumeServerAccessMode.equals("publicUrl")) {
+            filerClient.setAccessVolumeServerByPublicUrl();
+        } else if (volumeServerAccessMode.equals("filerProxy")) {
+            filerClient.setAccessVolumeServerByFilerProxy();
+        }
+
+    }
+
+    public void close() {
+        try {
+            this.filerClient.shutdown();
+        } catch (InterruptedException e) {
+            e.printStackTrace();
+        }
    }

    public static String getParentDirectory(Path path) {
@@ -185,6 +197,7 @@ 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);
@@ -204,10 +217,10 @@ public class SeaweedFileSystemStore {
                    .clearGroupName()
                    .addAllGroupName(Arrays.asList(userGroupInformation.getGroupNames()))
                );
-            SeaweedWrite.writeMeta(filerGrpcClient, getParentDirectory(path), entry);
+            SeaweedWrite.writeMeta(filerClient, getParentDirectory(path), entry);
        }

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

    }

@@ -221,7 +234,7 @@ public class SeaweedFileSystemStore {
            throw new FileNotFoundException("read non-exist file " + path);
        }

-        return new SeaweedInputStream(filerGrpcClient,
+        return new SeaweedHadoopInputStream(filerClient,
            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.FilerClient;
+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 FilerClient filerClient,
+            final Statistics statistics,
+            final String path,
+            final FilerProto.Entry entry) throws IOException {
+        this.seaweedInputStream = new SeaweedInputStream(filerClient, 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.FilerClient;
+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(FilerClient filerClient, final String path, FilerProto.Entry.Builder entry,
+                                     final long position, final int bufferSize, final String replication) {
+        super(filerClient, 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;
+        }
+    }
+
+}
@@ -1,338 +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.Arrays;
-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());
-            // 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
-}
+}
@@ -92,7 +92,7 @@ func main() {

 		header := superBlock.Bytes()

-		if n, e := datFile.WriteAt(header, 0); n == 0 || e != nil {
+		if n, e := datBackend.WriteAt(header, 0); n == 0 || e != nil {
 			glog.Fatalf("cannot write super block: %v", e)
 		}

@@ -168,7 +168,7 @@ func getVolumeFiles(v uint32, addr string) (map[types.NeedleId]needleState, int6
 				size:  size,
 			}
 		}
-		if actual := offset.ToAcutalOffset(); actual > maxOffset {
+		if actual := offset.ToActualOffset(); actual > maxOffset {
 			maxOffset = actual
 		}
 		return nil
@@ -0,0 +1,3 @@
+#!/bin/bash
+
+hsbench -a accesstoken -s secret -z 4K -d 10 -t 10 -b 10 -u http://localhost:8333 -m "cxipgdx" -bp "hsbench-"
@@ -45,7 +45,7 @@ func main() {
 			defer wg.Done()

 			client := &http.Client{Transport: &http.Transport{
-				MaxConnsPerHost:     1024,
+				MaxIdleConns:     1024,
 				MaxIdleConnsPerHost: 1024,
 			}}
 			r := rand.New(rand.NewSource(time.Now().UnixNano() + int64(x)))
@@ -0,0 +1,54 @@
+package main
+
+import (
+	"flag"
+	"fmt"
+	"math/rand"
+	"os"
+	"time"
+)
+
+var (
+	minSize   = flag.Int("minSize", 1024, "min file size")
+	maxSize   = flag.Int("maxSize", 3*1024*1024, "max file size")
+	fileCount = flag.Int("n", 1, "number of files to write")
+	blockSize = flag.Int("blockSizeKB", 4, "write block size")
+	toDir     = flag.String("dir", ".", "destination directory")
+)
+
+func check(e error) {
+	if e != nil {
+		panic(e)
+	}
+}
+
+func main() {
+
+	flag.Parse()
+
+	block := make([]byte, *blockSize*1024)
+
+	for i := 0; i < *fileCount; i++ {
+
+		f, err := os.Create(fmt.Sprintf("%s/file%05d", *toDir, i))
+		check(err)
+
+		fileSize := *minSize + rand.Intn(*maxSize-*minSize)
+		startTime := time.Now()
+
+		fmt.Printf("write %s %d bytes: ", f.Name(), fileSize)
+
+		for x := 0; x < fileSize; {
+			rand.Read(block)
+			_, err = f.Write(block)
+			check(err)
+			x += len(block)
+		}
+
+		err = f.Close()
+		check(err)
+
+		fmt.Printf("%.02f MB/sec\n", float64(fileSize)*float64(time.Second)/float64(time.Now().Sub(startTime)*1024*1024))
+	}
+
+}
@@ -37,7 +37,7 @@ func main() {
 		sinceTimeNs = time.Now().Add(-*rewindDuration).UnixNano()
 	}

-	err := operation.TailVolume(*master, grpcDialOption, vid, uint64(sinceTimeNs), *timeoutSeconds, func(n *needle.Needle) (err error) {
+	err := operation.TailVolume(func()string{return *master}, grpcDialOption, vid, uint64(sinceTimeNs), *timeoutSeconds, func(n *needle.Needle) (err error) {
 		if n.Size == 0 {
 			println("-", n.String())
 			return nil
@@ -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
@@ -72,7 +72,7 @@ func runBackup(cmd *Command, args []string) bool {
 	vid := needle.VolumeId(*s.volumeId)

 	// find volume location, replication, ttl info
-	lookup, err := operation.Lookup(*s.master, vid.String())
+	lookup, err := operation.Lookup(func() string { return *s.master }, vid.String())
 	if err != nil {
 		fmt.Printf("Error looking up volume %d: %v\n", vid, err)
 		return true
@@ -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
@@ -35,6 +35,7 @@ type BenchmarkOptions struct {
 	sequentialRead   *bool
 	collection       *string
 	replication      *string
+	diskType         *string
 	cpuprofile       *string
 	maxCpu           *int
 	grpcDialOption   grpc.DialOption
@@ -62,6 +63,7 @@ func init() {
 	b.sequentialRead = cmdBenchmark.Flag.Bool("readSequentially", false, "randomly read by ids from \"-list\" specified file")
 	b.collection = cmdBenchmark.Flag.String("collection", "benchmark", "write data to this collection")
 	b.replication = cmdBenchmark.Flag.String("replication", "000", "replication type")
+	b.diskType = cmdBenchmark.Flag.String("disk", "", "[hdd|ssd] hard drive or solid state drive")
 	b.cpuprofile = cmdBenchmark.Flag.String("cpuprofile", "", "cpu profile output file")
 	b.maxCpu = cmdBenchmark.Flag.Int("maxCpu", 0, "maximum number of CPUs. 0 means all available CPUs")
 	b.fsync = cmdBenchmark.Flag.Bool("fsync", false, "flush data to disk after write")
@@ -125,7 +127,7 @@ func runBenchmark(cmd *Command, args []string) bool {
 		defer pprof.StopCPUProfile()
 	}

-	b.masterClient = wdclient.NewMasterClient(b.grpcDialOption, "client", "", 0, strings.Split(*b.masters, ","))
+	b.masterClient = wdclient.NewMasterClient(b.grpcDialOption, "client", "", 0, "", strings.Split(*b.masters, ","))
 	go b.masterClient.KeepConnectedToMaster()
 	b.masterClient.WaitUntilConnected()

@@ -234,13 +236,14 @@ func writeFiles(idChan chan int, fileIdLineChan chan string, s *stat) {
 			Count:       1,
 			Collection:  *b.collection,
 			Replication: *b.replication,
+			DiskType:    *b.diskType,
 		}
-		if assignResult, err := operation.Assign(b.masterClient.GetMaster(), b.grpcDialOption, ar); err == nil {
+		if assignResult, err := operation.Assign(b.masterClient.GetMaster, b.grpcDialOption, ar); err == nil {
 			fp.Server, fp.Fid, fp.Collection = assignResult.Url, assignResult.Fid, *b.collection
 			if !isSecure && assignResult.Auth != "" {
 				isSecure = true
 			}
-			if _, err := fp.Upload(0, b.masterClient.GetMaster(), false, assignResult.Auth, b.grpcDialOption); err == nil {
+			if _, err := fp.Upload(0, b.masterClient.GetMaster, false, assignResult.Auth, b.grpcDialOption); err == nil {
 				if random.Intn(100) < *b.deletePercentage {
 					s.total++
 					delayedDeleteChan <- &delayedFile{time.Now().Add(time.Second), fp}
@@ -290,7 +293,7 @@ func readFiles(fileIdLineChan chan string, s *stat) {
 		}
 		var bytes []byte
 		for _, url := range urls {
-			bytes, _, err = util.Get(url)
+			bytes, _, err = util.FastGet(url)
 			if err == nil {
 				break
 			}
@@ -1,8 +1,8 @@
 package command

 import (
-	"flag"
 	"fmt"
+	flag "github.com/chrislusf/seaweedfs/weed/util/fla9"
 	"os"
 	"strings"
 )
@@ -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)
 	}
@@ -44,15 +44,15 @@ var cmdDownload = &Command{

 func runDownload(cmd *Command, args []string) bool {
 	for _, fid := range args {
-		if e := downloadToFile(*d.server, fid, util.ResolvePath(*d.dir)); e != nil {
+		if e := downloadToFile(func() string { return *d.server }, fid, util.ResolvePath(*d.dir)); e != nil {
 			fmt.Println("Download Error: ", fid, e)
 		}
 	}
 	return true
 }

-func downloadToFile(server, fileId, saveDir string) error {
-	fileUrl, lookupError := operation.LookupFileId(server, fileId)
+func downloadToFile(masterFn operation.GetMasterFn, fileId, saveDir string) error {
+	fileUrl, lookupError := operation.LookupFileId(masterFn, fileId)
 	if lookupError != nil {
 		return lookupError
 	}
@@ -83,7 +83,7 @@ func downloadToFile(server, fileId, saveDir string) error {
 		fids := strings.Split(string(content), "\n")
 		for _, partId := range fids {
 			var n int
-			_, part, err := fetchContent(*d.server, partId)
+			_, part, err := fetchContent(masterFn, partId)
 			if err == nil {
 				n, err = f.Write(part)
 			}
@@ -103,8 +103,8 @@ func downloadToFile(server, fileId, saveDir string) error {
 	return nil
 }

-func fetchContent(server string, fileId string) (filename string, content []byte, e error) {
-	fileUrl, lookupError := operation.LookupFileId(server, fileId)
+func fetchContent(masterFn operation.GetMasterFn, fileId string) (filename string, content []byte, e error) {
+	fileUrl, lookupError := operation.LookupFileId(masterFn, fileId)
 	if lookupError != nil {
 		return "", nil, lookupError
 	}
--- a/Show More
+++ b/Show More