docker move to go version 1.18

avoid invalid memory address or nil pointer dereference

.github/dependabot.yml

@@ -0,0 +1,10 @@
+version: 2
+updates:
+- package-ecosystem: "github-actions"
+  directory: "/"
+  schedule:
+      interval: "daily"
+- package-ecosystem: gomod
+  directory: "/"
+  schedule:
+    interval: daily

.github/workflows/binaries_dev.yml

@@ -4,15 +4,20 @@ on:
   push:
     branches: [ master ]
 
+permissions:
+  contents: read
+
 jobs:
 
   cleanup:
+    permissions:
+      contents: write  # for mknejp/delete-release-assets to delete release assets
     runs-on: ubuntu-latest
 
     steps:
 
       - name: Delete old release assets
-        uses: mknejp/delete-release-assets@v1
+        uses: mknejp/delete-release-assets@a8aaab13272b1eaac16cc46dddd3f725b97ee05a # v1
         with:
           token: ${{ github.token }}
           tag: dev
@@ -21,6 +26,8 @@ jobs:
             weed-*
 
   build_dev_linux_windows:
+    permissions:
+      contents: write  # for wangyoucao577/go-release-action to upload release assets
     needs: cleanup
     runs-on: ubuntu-latest
     strategy:
@@ -31,13 +38,13 @@ jobs:
     steps:
 
       - name: Check out code into the Go module directory
-        uses: actions/checkout@v2
+        uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
 
       - name: Set BUILD_TIME env
         run: echo BUILD_TIME=$(date -u +%Y%m%d-%H%M) >> ${GITHUB_ENV}
 
       - name: Go Release Binaries Large Disk
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}
@@ -53,7 +60,7 @@ jobs:
           asset_name: "weed-large-disk-${{ env.BUILD_TIME }}-${{ matrix.goos }}-${{ matrix.goarch }}"
 
       - name: Go Release Binaries Normal Volume Size
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}
@@ -68,6 +75,8 @@ jobs:
           asset_name: "weed-${{ env.BUILD_TIME }}-${{ matrix.goos }}-${{ matrix.goarch }}"
 
   build_dev_darwin:
+    permissions:
+      contents: write  # for wangyoucao577/go-release-action to upload release assets
     needs: build_dev_linux_windows
     runs-on: ubuntu-latest
     strategy:
@@ -78,13 +87,13 @@ jobs:
     steps:
 
       - name: Check out code into the Go module directory
-        uses: actions/checkout@v2
+        uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
 
       - name: Set BUILD_TIME env
         run: echo BUILD_TIME=$(date -u +%Y%m%d-%H%M) >> ${GITHUB_ENV}
 
       - name: Go Release Binaries Large Disk
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}
@@ -100,7 +109,7 @@ jobs:
           asset_name: "weed-large-disk-${{ env.BUILD_TIME }}-${{ matrix.goos }}-${{ matrix.goarch }}"
 
       - name: Go Release Binaries Normal Volume Size
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}

.github/workflows/binaries_release0.yml

@@ -11,9 +11,14 @@ on:
   workflow_dispatch:
 
 # A workflow run is made up of one or more jobs that can run sequentially or in parallel
+permissions:
+  contents: read
+
 jobs:
 
   build-release-binaries_windows:
+    permissions:
+      contents: write  # for wangyoucao577/go-release-action to upload release assets
     runs-on: ubuntu-latest
     strategy:
       matrix:
@@ -23,9 +28,9 @@ jobs:
     # Steps represent a sequence of tasks that will be executed as part of the job
     steps:
       # Checks-out your repository under $GITHUB_WORKSPACE, so your job can access it
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       - name: Go Release Binaries Normal Volume Size
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}
@@ -39,7 +44,7 @@ jobs:
           binary_name: weed
           asset_name: "${{ matrix.goos }}_${{ matrix.goarch }}"
       - name: Go Release Large Disk Binaries
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}

.github/workflows/binaries_release1.yml

@@ -11,9 +11,14 @@ on:
   workflow_dispatch:
 
 # A workflow run is made up of one or more jobs that can run sequentially or in parallel
+permissions:
+  contents: read
+
 jobs:
 
   build-release-binaries_linux:
+    permissions:
+      contents: write  # for wangyoucao577/go-release-action to upload release assets
     runs-on: ubuntu-latest
     strategy:
       matrix:
@@ -23,9 +28,9 @@ jobs:
     # Steps represent a sequence of tasks that will be executed as part of the job
     steps:
       # Checks-out your repository under $GITHUB_WORKSPACE, so your job can access it
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       - name: Go Release Binaries Normal Volume Size
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}
@@ -39,7 +44,7 @@ jobs:
           binary_name: weed
           asset_name: "${{ matrix.goos }}_${{ matrix.goarch }}"
       - name: Go Release Large Disk Binaries
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}

.github/workflows/binaries_release2.yml

@@ -11,9 +11,14 @@ on:
   workflow_dispatch:
 
 # A workflow run is made up of one or more jobs that can run sequentially or in parallel
+permissions:
+  contents: read
+
 jobs:
 
   build-release-binaries_darwin:
+    permissions:
+      contents: write  # for wangyoucao577/go-release-action to upload release assets
     runs-on: ubuntu-latest
     strategy:
       matrix:
@@ -23,9 +28,9 @@ jobs:
     # Steps represent a sequence of tasks that will be executed as part of the job
     steps:
       # Checks-out your repository under $GITHUB_WORKSPACE, so your job can access it
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       - name: Go Release Binaries Normal Volume Size
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}
@@ -39,7 +44,7 @@ jobs:
           binary_name: weed
           asset_name: "${{ matrix.goos }}_${{ matrix.goarch }}"
       - name: Go Release Large Disk Binaries
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}

.github/workflows/binaries_release3.yml

@@ -11,9 +11,14 @@ on:
   workflow_dispatch:
 
 # A workflow run is made up of one or more jobs that can run sequentially or in parallel
+permissions:
+  contents: read
+
 jobs:
 
   build-release-binaries_freebsd:
+    permissions:
+      contents: write  # for wangyoucao577/go-release-action to upload release assets
     runs-on: ubuntu-latest
     strategy:
       matrix:
@@ -23,9 +28,9 @@ jobs:
     # Steps represent a sequence of tasks that will be executed as part of the job
     steps:
       # Checks-out your repository under $GITHUB_WORKSPACE, so your job can access it
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       - name: Go Release Binaries Normal Volume Size
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}
@@ -39,7 +44,7 @@ jobs:
           binary_name: weed
           asset_name: "${{ matrix.goos }}_${{ matrix.goarch }}"
       - name: Go Release Large Disk Binaries
-        uses: wangyoucao577/go-release-action@v1.22
+        uses: wangyoucao577/go-release-action@16624612d4e2b73de613857a362d294700207fff # v1.22
         with:
           github_token: ${{ secrets.GITHUB_TOKEN }}
           goos: ${{ matrix.goos }}

.github/workflows/container_dev.yml

@@ -5,6 +5,9 @@ on:
     branches: [ master ]
   workflow_dispatch: []
 
+permissions:
+  contents: read
+
 jobs:
 
   build-dev-containers:
@@ -13,11 +16,11 @@ jobs:
     steps:
       -
         name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       -
         name: Docker meta
         id: docker_meta
-        uses: docker/metadata-action@v3
+        uses: docker/metadata-action@f2a13332ac1ce8c0a71aeac48a150dbb1838ab67 # v3
         with:
           images: |
             chrislusf/seaweedfs
@@ -30,30 +33,30 @@ jobs:
             org.opencontainers.image.vendor=Chris Lu
       -
         name: Set up QEMU
-        uses: docker/setup-qemu-action@v1
+        uses: docker/setup-qemu-action@27d0a4f181a40b142cce983c5393082c365d1480 # v1
       -
         name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v1
+        uses: docker/setup-buildx-action@94ab11c41e45d028884a99163086648e898eed25 # v1
         with:
           buildkitd-flags: "--debug"
       -
         name: Login to Docker Hub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v1
+        uses: docker/login-action@dd4fa0671be5250ee6f50aedf4cb05514abda2c7 # v1
         with:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
       -
         name: Login to GHCR
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v1
+        uses: docker/login-action@dd4fa0671be5250ee6f50aedf4cb05514abda2c7 # v1
         with:
           registry: ghcr.io
           username: ${{ secrets.GHCR_USERNAME }}
           password: ${{ secrets.GHCR_TOKEN }}
       -
         name: Build
-        uses: docker/build-push-action@v2
+        uses: docker/build-push-action@ac9327eae2b366085ac7f6a2d02df8aa8ead720a # v2
         with:
           context: ./docker
           push: ${{ github.event_name != 'pull_request' }}

.github/workflows/container_latest.yml

@@ -6,6 +6,9 @@ on:
       - '*'
   workflow_dispatch: []
 
+permissions:
+  contents: read
+
 jobs:
 
   build-latest-container:
@@ -14,11 +17,11 @@ jobs:
     steps:
       -
         name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       -
         name: Docker meta
         id: docker_meta
-        uses: docker/metadata-action@v3
+        uses: docker/metadata-action@f2a13332ac1ce8c0a71aeac48a150dbb1838ab67 # v3
         with:
           images: |
             chrislusf/seaweedfs
@@ -31,30 +34,30 @@ jobs:
             org.opencontainers.image.vendor=Chris Lu
       -
         name: Set up QEMU
-        uses: docker/setup-qemu-action@v1
+        uses: docker/setup-qemu-action@27d0a4f181a40b142cce983c5393082c365d1480 # v1
       -
         name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v1
+        uses: docker/setup-buildx-action@94ab11c41e45d028884a99163086648e898eed25 # v1
         with:
           buildkitd-flags: "--debug"
       -
         name: Login to Docker Hub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v1
+        uses: docker/login-action@dd4fa0671be5250ee6f50aedf4cb05514abda2c7 # v1
         with:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
       -
         name: Login to GHCR
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v1
+        uses: docker/login-action@dd4fa0671be5250ee6f50aedf4cb05514abda2c7 # v1
         with:
           registry: ghcr.io
           username: ${{ secrets.GHCR_USERNAME }}
           password: ${{ secrets.GHCR_TOKEN }}
       -
         name: Build
-        uses: docker/build-push-action@v2
+        uses: docker/build-push-action@ac9327eae2b366085ac7f6a2d02df8aa8ead720a # v2
         with:
           context: ./docker
           push: ${{ github.event_name != 'pull_request' }}

.github/workflows/container_release1.yml

@@ -6,6 +6,9 @@ on:
       - '*'
   workflow_dispatch: []
 
+permissions:
+  contents: read
+
 jobs:
   build-default-release-container:
     runs-on: [ubuntu-latest]
@@ -13,11 +16,11 @@ jobs:
     steps:
       -
         name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       -
         name: Docker meta
         id: docker_meta
-        uses: docker/metadata-action@v3
+        uses: docker/metadata-action@f2a13332ac1ce8c0a71aeac48a150dbb1838ab67 # v3
         with:
           images: |
             chrislusf/seaweedfs
@@ -31,20 +34,20 @@ jobs:
             org.opencontainers.image.vendor=Chris Lu
       -
         name: Set up QEMU
-        uses: docker/setup-qemu-action@v1
+        uses: docker/setup-qemu-action@27d0a4f181a40b142cce983c5393082c365d1480 # v1
       -
         name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v1
+        uses: docker/setup-buildx-action@94ab11c41e45d028884a99163086648e898eed25 # v1
       -
         name: Login to Docker Hub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v1
+        uses: docker/login-action@dd4fa0671be5250ee6f50aedf4cb05514abda2c7 # v1
         with:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
       -
         name: Build
-        uses: docker/build-push-action@v2
+        uses: docker/build-push-action@ac9327eae2b366085ac7f6a2d02df8aa8ead720a # v2
         with:
           context: ./docker
           push: ${{ github.event_name != 'pull_request' }}

.github/workflows/container_release2.yml

@@ -6,6 +6,9 @@ on:
       - '*'
   workflow_dispatch: []
 
+permissions:
+  contents: read
+
 jobs:
 
   build-large-release-container:
@@ -14,11 +17,11 @@ jobs:
     steps:
       -
         name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       -
         name: Docker meta
         id: docker_meta
-        uses: docker/metadata-action@v3
+        uses: docker/metadata-action@f2a13332ac1ce8c0a71aeac48a150dbb1838ab67 # v3
         with:
           images: |
             chrislusf/seaweedfs
@@ -32,20 +35,20 @@ jobs:
             org.opencontainers.image.vendor=Chris Lu
       -
         name: Set up QEMU
-        uses: docker/setup-qemu-action@v1
+        uses: docker/setup-qemu-action@27d0a4f181a40b142cce983c5393082c365d1480 # v1
       -
         name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v1
+        uses: docker/setup-buildx-action@94ab11c41e45d028884a99163086648e898eed25 # v1
       -
         name: Login to Docker Hub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v1
+        uses: docker/login-action@dd4fa0671be5250ee6f50aedf4cb05514abda2c7 # v1
         with:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
       -
         name: Build
-        uses: docker/build-push-action@v2
+        uses: docker/build-push-action@ac9327eae2b366085ac7f6a2d02df8aa8ead720a # v2
         with:
           context: ./docker
           push: ${{ github.event_name != 'pull_request' }}

.github/workflows/container_release3.yml

@@ -6,6 +6,9 @@ on:
       - '*'
   workflow_dispatch: []
 
+permissions:
+  contents: read
+
 jobs:
 
   build-large-release-container_rocksdb:
@@ -14,11 +17,11 @@ jobs:
     steps:
       -
         name: Checkout
-        uses: actions/checkout@v2
+        uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
       -
         name: Docker meta
         id: docker_meta
-        uses: docker/metadata-action@v3
+        uses: docker/metadata-action@f2a13332ac1ce8c0a71aeac48a150dbb1838ab67 # v3
         with:
           images: |
             chrislusf/seaweedfs
@@ -32,20 +35,20 @@ jobs:
             org.opencontainers.image.vendor=Chris Lu
       -
         name: Set up QEMU
-        uses: docker/setup-qemu-action@v1
+        uses: docker/setup-qemu-action@27d0a4f181a40b142cce983c5393082c365d1480 # v1
       -
         name: Set up Docker Buildx
-        uses: docker/setup-buildx-action@v1
+        uses: docker/setup-buildx-action@94ab11c41e45d028884a99163086648e898eed25 # v1
       -
         name: Login to Docker Hub
         if: github.event_name != 'pull_request'
-        uses: docker/login-action@v1
+        uses: docker/login-action@dd4fa0671be5250ee6f50aedf4cb05514abda2c7 # v1
         with:
           username: ${{ secrets.DOCKER_USERNAME }}
           password: ${{ secrets.DOCKER_PASSWORD }}
       -
         name: Build
-        uses: docker/build-push-action@v2
+        uses: docker/build-push-action@ac9327eae2b366085ac7f6a2d02df8aa8ead720a # v2
         with:
           context: ./docker
           push: ${{ github.event_name != 'pull_request' }}

.github/workflows/go.yml

@@ -10,6 +10,9 @@ concurrency:
   group: ${{ github.head_ref }}/go
   cancel-in-progress: true
 
+permissions:
+  contents: read
+
 jobs:
 
   build:
@@ -18,13 +21,13 @@ jobs:
     steps:
 
     - name: Set up Go 1.x
-      uses: actions/setup-go@v2
+      uses: actions/setup-go@f6164bd8c8acb4a71fb2791a8b6c4024ff038dab # v2
       with:
         go-version: ^1.13
       id: go
 
     - name: Check out code into the Go module directory
-      uses: actions/checkout@v2
+      uses: actions/checkout@629c2de402a417ea7690ca6ce3f33229e27606a5 # v2
 
     - name: Get dependencies
       run: |

README.md

@@ -33,7 +33,7 @@ Your support will be really appreciated by me and other supporters!
 
 
 ### Gold Sponsors
-![shuguang](https://raw.githubusercontent.com/chrislusf/seaweedfs/master/note/shuguang.png)
+- [![nodion](https://www.nodion.com/img/logo.svg)](https://www.nodion.com)
 
 ---
 
@@ -52,6 +52,9 @@ Table of Contents
 =================
 
 * [Quick Start](#quick-start)
+    * [Quick Start for S3 API on Docker](#quick-start-for-s3-api-on-docker)
+    * [Quick Start with Single Binary](#quick-start-with-single-binary)
+    * [Quick Start SeaweedFS S3 on AWS](#quick-start-seaweedfs-s3-on-aws)
 * [Introduction](#introduction)
 * [Features](#features)
     * [Additional Features](#additional-features)
@@ -74,12 +77,15 @@ Table of Contents
 
 `docker run -p 8333:8333 chrislusf/seaweedfs server -s3`
 
-## Quick Start with single binary ##
+## Quick Start with Single Binary ##
 * 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 thousands of machines. That is it!
 
+## Quick Start SeaweedFS S3 on AWS ##
+* Setup fast production-ready [SeaweedFS S3 on AWS with cloudformation](https://aws.amazon.com/marketplace/pp/prodview-nzelz5gprlrjc)
+
 ## Introduction ##
 
 SeaweedFS is a simple and highly scalable distributed file system. There are two objectives:

docker/Dockerfile.go_build

@@ -1,4 +1,4 @@
-FROM golang:1.17-alpine as builder
+FROM golang:1.18-alpine as builder
 RUN apk add git g++ fuse
 RUN mkdir -p /go/src/github.com/chrislusf/
 RUN git clone https://github.com/chrislusf/seaweedfs /go/src/github.com/chrislusf/seaweedfs

docker/Dockerfile.go_build_large

@@ -1,4 +1,4 @@
-FROM golang:1.17-alpine as builder
+FROM golang:1.18-alpine as builder
 RUN apk add git g++ fuse
 RUN mkdir -p /go/src/github.com/chrislusf/
 RUN git clone https://github.com/chrislusf/seaweedfs /go/src/github.com/chrislusf/seaweedfs

docker/Dockerfile.rocksdb_large

@@ -3,7 +3,7 @@ FROM golang:1.17-buster as builder
 RUN apt-get update
 RUN apt-get install -y build-essential libsnappy-dev zlib1g-dev libbz2-dev libgflags-dev liblz4-dev libzstd-dev
 
-ENV ROCKSDB_VERSION v6.22.1
+ENV ROCKSDB_VERSION v7.0.4
 
 # build RocksDB
 RUN cd /tmp && \

docker/Makefile

@@ -55,6 +55,9 @@ cluster: build
 2clusters: build
 	docker-compose -f compose/local-clusters-compose.yml -p seaweedfs up
 
+hashicorp_raft: build
+	docker-compose -f compose/local-hashicorp-raft-compose.yml -p seaweedfs up
+
 s3tests: build s3tests_build
 	docker-compose -f compose/local-s3tests-compose.yml -p seaweedfs up
 

docker/compose/local-cluster-compose.yml

@@ -36,7 +36,7 @@ services:
     ports:
       - 8080:8080
       - 18080:18080
-    command: 'volume -dataCenter dc1 -rack v1 -mserver="master0:9333,master1:9334,master2:9335" -port=8080 -ip=volume1 -publicUrl=localhost:8080 -preStopSeconds=1'
+    command: 'volume -dataCenter=dc1 -rack=v1 -mserver="master0:9333,master1:9334,master2:9335" -port=8080 -ip=volume1 -publicUrl=localhost:8080 -preStopSeconds=1'
     depends_on:
       - master0
       - master1
@@ -46,7 +46,7 @@ services:
     ports:
       - 8082:8082
       - 18082:18082
-    command: 'volume -dataCenter dc2 -rack v2 -mserver="master0:9333,master1:9334,master2:9335" -port=8082 -ip=volume2 -publicUrl=localhost:8082 -preStopSeconds=1'
+    command: 'volume -dataCenter=dc2 -rack=v2 -mserver="master0:9333,master1:9334,master2:9335" -port=8082 -ip=volume2 -publicUrl=localhost:8082 -preStopSeconds=1'
     depends_on:
       - master0
       - master1
@@ -56,7 +56,7 @@ services:
     ports:
       - 8083:8083
       - 18083:18083
-    command: 'volume -dataCenter dc3 -rack v3 -mserver="master0:9333,master1:9334,master2:9335" -port=8083 -ip=volume3 -publicUrl=localhost:8083 -preStopSeconds=1'
+    command: 'volume -dataCenter=dc3 -rack=v3 -mserver="master0:9333,master1:9334,master2:9335" -port=8083 -ip=volume3 -publicUrl=localhost:8083 -preStopSeconds=1'
     depends_on:
       - master0
       - master1
@@ -67,7 +67,7 @@ services:
       - 8888:8888
       - 18888:18888
       - 8111:8111
-    command: 'filer -defaultReplicaPlacement 100 -iam -master="master0:9333,master1:9334,master2:9335"'
+    command: 'filer -defaultReplicaPlacement=100 -iam -master="master0:9333,master1:9334,master2:9335"'
     depends_on:
       - master0
       - master1
@@ -85,4 +85,4 @@ services:
       - master2
       - volume1
       - volume2
-      - filer
+      - filer

docker/compose/local-hashicorp-raft-compose.yml

@@ -0,0 +1,89 @@
+version: '2'
+
+services:
+  master0:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9333:9333
+      - 19333:19333
+    command: "-v=4 master -volumeSizeLimitMB 100 -raftHashicorp -ip=master0 -port=9333 -peers=master1:9334,master2:9335 -mdir=/data"
+    volumes:
+      - ./master/0:/data
+    environment:
+      WEED_MASTER_VOLUME_GROWTH_COPY_1: 1
+      WEED_MASTER_VOLUME_GROWTH_COPY_2: 2
+      WEED_MASTER_VOLUME_GROWTH_COPY_OTHER: 1
+  master1:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9334:9334
+      - 19334:19334
+    command: "-v=4 master -volumeSizeLimitMB 100 -raftHashicorp -ip=master1 -port=9334 -peers=master0:9333,master2:9335 -mdir=/data"
+    volumes:
+      - ./master/1:/data
+    environment:
+      WEED_MASTER_VOLUME_GROWTH_COPY_1: 1
+      WEED_MASTER_VOLUME_GROWTH_COPY_2: 2
+      WEED_MASTER_VOLUME_GROWTH_COPY_OTHER: 1
+  master2:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 9335:9335
+      - 19335:19335
+    command: "-v=4 master -volumeSizeLimitMB 100 -raftHashicorp -ip=master2 -port=9335 -peers=master0:9333,master1:9334 -mdir=/data"
+    volumes:
+      - ./master/2:/data
+    environment:
+      WEED_MASTER_VOLUME_GROWTH_COPY_1: 1
+      WEED_MASTER_VOLUME_GROWTH_COPY_2: 2
+      WEED_MASTER_VOLUME_GROWTH_COPY_OTHER: 1
+  volume1:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8080:8080
+      - 18080:18080
+    command: 'volume -dataCenter=dc1 -rack=v1 -mserver="master0:9333,master1:9334,master2:9335" -port=8080 -ip=volume1 -publicUrl=localhost:8080 -preStopSeconds=1'
+    depends_on:
+      - master0
+      - master1
+  volume2:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8082:8082
+      - 18082:18082
+    command: 'volume -dataCenter=dc2 -rack=v2 -mserver="master0:9333,master1:9334,master2:9335" -port=8082 -ip=volume2 -publicUrl=localhost:8082 -preStopSeconds=1'
+    depends_on:
+      - master0
+      - master1
+  volume3:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8083:8083
+      - 18083:18083
+    command: 'volume -dataCenter=dc3 -rack=v3 -mserver="master0:9333,master1:9334,master2:9335" -port=8083 -ip=volume3 -publicUrl=localhost:8083 -preStopSeconds=1'
+    depends_on:
+      - master0
+      - master1
+  filer:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8888:8888
+      - 18888:18888
+      - 8111:8111
+    command: 'filer -defaultReplicaPlacement=100 -iam -master="master0:9333,master1:9334,master2:9335"'
+    depends_on:
+      - master0
+      - master1
+      - volume1
+      - volume2
+  s3:
+    image: chrislusf/seaweedfs:local
+    ports:
+      - 8333:8333
+    command: '-v=9 s3 -ip.bind="s3" -filer="filer:8888"'
+    depends_on:
+      - master0
+      - master1
+      - volume1
+      - volume2
+      - filer

docker/compose/local-s3tests-compose.yml

@@ -24,7 +24,7 @@ services:
       - 8888:8888
       - 18888:18888
       - 8000:8000
-    command: 'filer -master="master:9333" -s3 -s3.config=/etc/seaweedfs/s3.json -s3.port=8000'
+    command: 'filer -master="master:9333" -s3 -s3.config=/etc/seaweedfs/s3.json -s3.port=8000 -s3.allowEmptyFolder=false -s3.allowDeleteBucketNotEmpty=false'
     volumes:
       - ./s3.json:/etc/seaweedfs/s3.json
     depends_on:
@@ -38,7 +38,7 @@ services:
       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: "(get_bucket_encryption|put_bucket_encryption|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_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_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_ifunmodifiedsince_good|put_object_ifmatch_failed|object_raw_get_bucket_gone|object_delete_key_bucket_gone|object_raw_get_bucket_acl|object_raw_get_object_acl|object_raw_response_headers|object_raw_authenticated_bucket_gone|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_exists|bucket_create_naming_bad_short_one|bucket_create_naming_bad_short_two|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|object_copy_verify_contenttype|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)"
+      NOSETESTS_EXCLUDE: "(get_bucket_encryption|put_bucket_encryption|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_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_concurrent_set_canned_acl|object_write_to_nonexist_bucket|object_requestid_matches_header_on_error|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_ifunmodifiedsince_good|put_object_ifmatch_failed|object_raw_get_bucket_gone|object_delete_key_bucket_gone|object_raw_get_bucket_acl|object_raw_get_object_acl|object_raw_response_headers|object_raw_authenticated_bucket_gone|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_exists|bucket_create_naming_bad_short_one|bucket_create_naming_bad_short_two|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|object_copy_verify_contenttype|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|100_continue|ranged_request_invalid_range|ranged_request_empty_object|access_bucket|list_multipart_upload_owner|multipart_upload_small)"
     depends_on:
       - master
       - volume

docker/entrypoint.sh

@@ -24,7 +24,7 @@ case "$1" in
   'master')
   	ARGS="-mdir=/data -volumePreallocate -volumeSizeLimitMB=1024"
   	shift
-  	exec /usr/bin/weed master $ARGS $@
+  	exec /usr/bin/weed -logtostderr=true master $ARGS $@
 	;;
 
   'volume')
@@ -33,7 +33,7 @@ case "$1" in
   	  ARGS="-dir=/data"
   	fi
   	shift
-  	exec /usr/bin/weed volume $ARGS $@
+  	exec /usr/bin/weed -logtostderr=true volume $ARGS $@
 	;;
 
   'server')
@@ -42,19 +42,19 @@ case "$1" in
   	  ARGS="-dir=/data -master.volumePreallocate -master.volumeSizeLimitMB=1024"
   	fi
  	shift
-  	exec /usr/bin/weed server $ARGS $@
+  	exec /usr/bin/weed -logtostderr=true server $ARGS $@
   	;;
 
   'filer')
   	ARGS=""
   	shift
-  	exec /usr/bin/weed filer $ARGS $@
+  	exec /usr/bin/weed -logtostderr=true filer $ARGS $@
 	;;
 
   's3')
   	ARGS="-domainName=$S3_DOMAIN_NAME -key.file=$S3_KEY_FILE -cert.file=$S3_CERT_FILE"
   	shift
-  	exec /usr/bin/weed s3 $ARGS $@
+  	exec /usr/bin/weed -logtostderr=true s3 $ARGS $@
 	;;
 
   *)

go.mod

@@ -1,23 +1,23 @@
 module github.com/chrislusf/seaweedfs
 
-go 1.17
+go 1.18
 
 require (
-	cloud.google.com/go v0.94.1 // indirect
-	cloud.google.com/go/pubsub v1.3.1
-	cloud.google.com/go/storage v1.16.1
+	cloud.google.com/go v0.100.2 // indirect
+	cloud.google.com/go/pubsub v1.20.0
+	cloud.google.com/go/storage v1.22.0
 	github.com/Azure/azure-pipeline-go v0.2.3
 	github.com/Azure/azure-storage-blob-go v0.14.0
-	github.com/OneOfOne/xxhash v1.2.2
-	github.com/Shopify/sarama v1.23.1
-	github.com/aws/aws-sdk-go v1.35.3
+	github.com/OneOfOne/xxhash v1.2.8
+	github.com/Shopify/sarama v1.32.0
+	github.com/aws/aws-sdk-go v1.43.40
 	github.com/beorn7/perks v1.0.1 // indirect
 	github.com/buraksezer/consistent v0.0.0-20191006190839-693edf70fd72
 	github.com/bwmarrin/snowflake v0.3.0
 	github.com/cespare/xxhash v1.1.0
-	github.com/cespare/xxhash/v2 v2.1.1 // indirect
-	github.com/chrislusf/raft v1.0.7
-	github.com/colinmarc/hdfs/v2 v2.2.0
+	github.com/cespare/xxhash/v2 v2.1.2 // indirect
+	github.com/chrislusf/raft v1.0.8
+	github.com/colinmarc/hdfs/v2 v2.3.0
 	github.com/coreos/go-semver v0.3.0 // indirect
 	github.com/coreos/go-systemd/v22 v22.3.2 // indirect
 	github.com/davecgh/go-spew v1.1.1 // indirect
@@ -32,86 +32,77 @@ 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/fsnotify/fsnotify v1.4.9 // indirect
+	github.com/fclairamb/ftpserverlib v0.17.0
+	github.com/fsnotify/fsnotify v1.5.1 // indirect
 	github.com/go-errors/errors v1.1.1 // indirect
-	github.com/go-redis/redis/v8 v8.4.4
-	github.com/go-redsync/redsync/v4 v4.4.1
-	github.com/go-sql-driver/mysql v1.5.0
+	github.com/go-redis/redis/v8 v8.11.5
+	github.com/go-redsync/redsync/v4 v4.5.0
+	github.com/go-sql-driver/mysql v1.6.0
 	github.com/go-stack/stack v1.8.0 // indirect
 	github.com/go-zookeeper/zk v1.0.2 // indirect
 	github.com/gocql/gocql v0.0.0-20210707082121-9a3953d1826d
-	github.com/golang-jwt/jwt v3.2.1+incompatible
-	github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da
+	github.com/golang-jwt/jwt v3.2.2+incompatible
+	github.com/golang/groupcache v0.0.0-20210331224755-41bb18bfe9da // indirect
 	github.com/golang/protobuf v1.5.2
 	github.com/golang/snappy v0.0.4 // indirect
-	github.com/google/btree v1.0.0
-	github.com/google/go-cmp v0.5.6 // indirect
+	github.com/google/btree v1.0.1
+	github.com/google/go-cmp v0.5.7 // indirect
 	github.com/google/uuid v1.3.0
-	github.com/google/wire v0.4.0 // indirect
-	github.com/googleapis/gax-go v2.0.2+incompatible // indirect
-	github.com/googleapis/gax-go/v2 v2.1.0 // indirect
-	github.com/gorilla/mux v1.7.4
-	github.com/grpc-ecosystem/go-grpc-middleware v1.1.0
+	github.com/google/wire v0.5.0 // indirect
+	github.com/googleapis/gax-go/v2 v2.3.0 // indirect
+	github.com/gorilla/mux v1.8.0
+	github.com/grpc-ecosystem/go-grpc-middleware v1.3.0
 	github.com/hailocab/go-hostpool v0.0.0-20160125115350-e80d13ce29ed // indirect
 	github.com/hashicorp/errwrap v1.0.0 // indirect
 	github.com/hashicorp/go-multierror v1.1.0 // indirect
 	github.com/hashicorp/go-uuid v1.0.2 // indirect
 	github.com/hashicorp/hcl v1.0.0 // indirect
 	github.com/jcmturner/gofork v1.0.0 // indirect
-	github.com/jcmturner/gokrb5/v8 v8.4.1
-	github.com/jinzhu/copier v0.2.8
+	github.com/jcmturner/gokrb5/v8 v8.4.2
+	github.com/jinzhu/copier v0.3.5
 	github.com/jmespath/go-jmespath v0.4.0 // indirect
-	github.com/json-iterator/go v1.1.11
-	github.com/karlseguin/ccache/v2 v2.0.7
+	github.com/json-iterator/go v1.1.12
+	github.com/karlseguin/ccache/v2 v2.0.8
 	github.com/kballard/go-shellquote v0.0.0-20180428030007-95032a82bc51 // indirect
-	github.com/klauspost/compress v1.13.6 // indirect
-	github.com/klauspost/cpuid v1.2.1 // indirect
-	github.com/klauspost/reedsolomon v1.9.2
-	github.com/konsorten/go-windows-terminal-sequences v1.0.3 // indirect
+	github.com/klauspost/compress v1.15.1 // indirect
+	github.com/klauspost/reedsolomon v1.9.16
 	github.com/kurin/blazer v0.5.3
-	github.com/lib/pq v1.10.0
-	github.com/linxGnu/grocksdb v1.6.38
-	github.com/magiconair/properties v1.8.1 // indirect
-	github.com/mailru/easyjson v0.7.1 // indirect
-	github.com/mattn/go-ieproxy v0.0.1 // indirect
-	github.com/mattn/go-isatty v0.0.12 // indirect
+	github.com/lib/pq v1.10.5
+	github.com/linxGnu/grocksdb v1.7.0
+	github.com/magiconair/properties v1.8.6 // indirect
+	github.com/mailru/easyjson v0.7.7 // indirect
+	github.com/mattn/go-ieproxy v0.0.3 // indirect
+	github.com/mattn/go-isatty v0.0.14 // indirect
 	github.com/matttproud/golang_protobuf_extensions v1.0.1 // indirect
-	github.com/mitchellh/mapstructure v1.1.2 // indirect
+	github.com/mitchellh/mapstructure v1.4.3 // indirect
 	github.com/modern-go/concurrent v0.0.0-20180306012644-bacd9c7ef1dd // indirect
-	github.com/modern-go/reflect2 v1.0.1 // indirect
-	github.com/nats-io/jwt v1.0.1 // indirect
-	github.com/nats-io/nats.go v1.10.0 // indirect
-	github.com/nats-io/nkeys v0.2.0 // indirect
-	github.com/nats-io/nuid v1.0.1 // indirect
-	github.com/olivere/elastic/v7 v7.0.19
-	github.com/pelletier/go-toml v1.7.0 // indirect
-	github.com/peterh/liner v1.1.0
-	github.com/pierrec/lz4 v2.2.7+incompatible // indirect
+	github.com/modern-go/reflect2 v1.0.2 // indirect
+	github.com/olivere/elastic/v7 v7.0.32
+	github.com/pelletier/go-toml v1.9.4 // indirect
+	github.com/peterh/liner v1.2.2
+	github.com/pierrec/lz4 v2.6.1+incompatible // indirect
 	github.com/pkg/errors v0.9.1 // indirect
 	github.com/pmezard/go-difflib v1.0.0 // indirect
 	github.com/posener/complete v1.2.3
 	github.com/pquerna/cachecontrol v0.1.0
-	github.com/prometheus/client_golang v1.11.0
+	github.com/prometheus/client_golang v1.12.1
 	github.com/prometheus/client_model v0.2.0 // indirect
-	github.com/prometheus/common v0.26.0 // indirect
-	github.com/prometheus/procfs v0.6.0 // indirect
-	github.com/rcrowley/go-metrics v0.0.0-20190826022208-cac0b30c2563 // indirect
+	github.com/prometheus/common v0.32.1 // indirect
+	github.com/prometheus/procfs v0.7.3 // indirect
+	github.com/rcrowley/go-metrics v0.0.0-20201227073835-cf1acfcdf475 // indirect
 	github.com/remyoudompheng/bigfft v0.0.0-20200410134404-eec4a21b6bb0 // indirect
-	github.com/seaweedfs/fuse v1.2.2
-	github.com/seaweedfs/goexif v1.0.2
-	github.com/sirupsen/logrus v1.6.0 // indirect
+	github.com/seaweedfs/goexif v2.0.0+incompatible
+	github.com/sirupsen/logrus v1.8.1 // indirect
 	github.com/spaolacci/murmur3 v1.1.0 // indirect
-	github.com/spf13/afero v1.6.0 // indirect
-	github.com/spf13/cast v1.3.0 // indirect
+	github.com/spf13/afero v1.8.2 // indirect
+	github.com/spf13/cast v1.4.1 // indirect
 	github.com/spf13/jwalterweatherman v1.1.0 // indirect
-	github.com/spf13/viper v1.4.0
-	github.com/streadway/amqp v0.0.0-20200108173154-1c71cc93ed71
-	github.com/stretchr/testify v1.7.0
+	github.com/spf13/viper v1.11.0
+	github.com/streadway/amqp v1.0.0
+	github.com/stretchr/testify v1.7.1
 	github.com/stvp/tempredis v0.0.0-20181119212430-b82af8480203
 	github.com/syndtr/goleveldb v1.0.1-0.20190318030020-c3a204f8e965
-	github.com/tidwall/gjson v1.10.2
+	github.com/tidwall/gjson v1.14.0
 	github.com/tidwall/match v1.1.1
 	github.com/tidwall/pretty v1.2.0 // indirect
 	github.com/tsuna/gohbase v0.0.0-20201125011725-348991136365
@@ -121,76 +112,106 @@ require (
 	github.com/viant/ptrie v0.3.0
 	github.com/viant/toolbox v0.33.2 // indirect
 	github.com/xdg-go/pbkdf2 v1.0.0 // indirect
-	github.com/xdg-go/scram v1.0.2 // indirect
+	github.com/xdg-go/scram v1.1.0 // indirect
 	github.com/xdg-go/stringprep v1.0.2 // indirect
 	github.com/youmark/pkcs8 v0.0.0-20181117223130-1be2e3e5546d // indirect
-	go.etcd.io/etcd/client/v3 v3.5.0
-	go.mongodb.org/mongo-driver v1.8.0
+	go.etcd.io/etcd/client/v3 v3.5.3
+	go.mongodb.org/mongo-driver v1.9.0
 	go.opencensus.io v0.23.0 // indirect
-	go.opentelemetry.io/otel v0.15.0 // indirect
-	gocloud.dev v0.20.0
-	gocloud.dev/pubsub/natspubsub v0.20.0
-	gocloud.dev/pubsub/rabbitpubsub v0.20.0
-	golang.org/x/crypto v0.0.0-20201216223049-8b5274cf687f // indirect
+	gocloud.dev v0.25.0
+	gocloud.dev/pubsub/natspubsub v0.25.0
+	gocloud.dev/pubsub/rabbitpubsub v0.25.0
+	golang.org/x/crypto v0.0.0-20220411220226-7b82a4e95df4 // indirect
+	golang.org/x/exp v0.0.0-20220414153411-bcd21879b8fd
 	golang.org/x/image v0.0.0-20200119044424-58c23975cae1
-	golang.org/x/net v0.0.0-20210813160813-60bc85c4be6d
-	golang.org/x/oauth2 v0.0.0-20210819190943-2bc19b11175f // indirect
-	golang.org/x/sys v0.0.0-20220111092808-5a964db01320
+	golang.org/x/net v0.0.0-20220412020605-290c469a71a5
+	golang.org/x/oauth2 v0.0.0-20220411215720-9780585627b5 // indirect
+	golang.org/x/sys v0.0.0-20220412211240-33da011f77ad
 	golang.org/x/text v0.3.7 // indirect
-	golang.org/x/tools v0.1.6-0.20210726203631-07bc1bf47fb2
-	golang.org/x/xerrors v0.0.0-20200804184101-5ec99f83aff1 // indirect
-	google.golang.org/api v0.57.0
+	golang.org/x/tools v0.1.8-0.20211029000441-d6a9af8af023
+	golang.org/x/xerrors v0.0.0-20220411194840-2f41105eb62f // indirect
+	google.golang.org/api v0.74.0
 	google.golang.org/appengine v1.6.7 // indirect
-	google.golang.org/genproto v0.0.0-20210921142501-181ce0d877f6 // indirect
-	google.golang.org/grpc v1.40.0
-	google.golang.org/protobuf v1.27.1
+	google.golang.org/genproto v0.0.0-20220407144326-9054f6ed7bac // indirect
+	google.golang.org/grpc v1.45.0
+	google.golang.org/protobuf v1.28.0
 	gopkg.in/inf.v0 v0.9.1 // indirect
-	gopkg.in/jcmturner/aescts.v1 v1.0.1 // indirect
-	gopkg.in/jcmturner/dnsutils.v1 v1.0.1 // indirect
-	gopkg.in/jcmturner/goidentity.v3 v3.0.0 // indirect
-	gopkg.in/jcmturner/gokrb5.v7 v7.3.0 // indirect
-	gopkg.in/jcmturner/rpc.v1 v1.1.0 // indirect
 	modernc.org/b v1.0.0 // indirect
-	modernc.org/cc/v3 v3.33.5 // indirect
-	modernc.org/ccgo/v3 v3.9.4 // indirect
-	modernc.org/libc v1.9.5 // indirect
-	modernc.org/mathutil v1.2.2 // indirect
-	modernc.org/memory v1.0.4 // indirect
+	modernc.org/cc/v3 v3.35.24 // indirect
+	modernc.org/ccgo/v3 v3.15.18 // indirect
+	modernc.org/libc v1.14.12 // indirect
+	modernc.org/mathutil v1.4.1 // indirect
+	modernc.org/memory v1.0.7 // indirect
 	modernc.org/opt v0.1.1 // indirect
-	modernc.org/sqlite v1.10.7
-	modernc.org/strutil v1.1.0 // indirect
+	modernc.org/sqlite v1.16.0
+	modernc.org/strutil v1.1.1 // indirect
 	modernc.org/token v1.0.0 // indirect
 )
 
 require (
-	github.com/fluent/fluent-logger-golang v1.8.0
+	github.com/Jille/raft-grpc-transport v1.2.0
+	github.com/fluent/fluent-logger-golang v1.9.0
 	github.com/hanwen/go-fuse/v2 v2.1.0
+	github.com/hashicorp/raft v1.3.7
+	github.com/hashicorp/raft-boltdb v0.0.0-20220329195025-15018e9b97e0
 )
 
 require (
-	cloud.google.com/go/kms v1.0.0 // indirect
-	github.com/DataDog/zstd v1.3.6-0.20190409195224-796139022798 // indirect
+	cloud.google.com/go/compute v1.5.0 // indirect
+	cloud.google.com/go/iam v0.3.0 // indirect
+	github.com/armon/go-metrics v0.3.10 // indirect
+	github.com/aws/aws-sdk-go-v2 v1.16.2 // indirect
+	github.com/aws/aws-sdk-go-v2/config v1.15.3 // indirect
+	github.com/aws/aws-sdk-go-v2/credentials v1.11.2 // indirect
+	github.com/aws/aws-sdk-go-v2/feature/ec2/imds v1.12.3 // indirect
+	github.com/aws/aws-sdk-go-v2/internal/configsources v1.1.9 // indirect
+	github.com/aws/aws-sdk-go-v2/internal/endpoints/v2 v2.4.3 // indirect
+	github.com/aws/aws-sdk-go-v2/internal/ini v1.3.10 // indirect
+	github.com/aws/aws-sdk-go-v2/service/internal/presigned-url v1.9.3 // indirect
+	github.com/aws/aws-sdk-go-v2/service/sns v1.17.4 // indirect
+	github.com/aws/aws-sdk-go-v2/service/sqs v1.18.3 // indirect
+	github.com/aws/aws-sdk-go-v2/service/sso v1.11.3 // indirect
+	github.com/aws/aws-sdk-go-v2/service/sts v1.16.3 // indirect
+	github.com/aws/smithy-go v1.11.2 // indirect
+	github.com/boltdb/bolt v1.3.1 // indirect
 	github.com/d4l3k/messagediff v1.2.1 // indirect
+	github.com/fatih/color v1.13.0 // indirect
+	github.com/fclairamb/go-log v0.1.0 // indirect
 	github.com/gogo/protobuf v1.3.2 // indirect
+	github.com/googleapis/go-type-adapters v1.0.0 // indirect
+	github.com/hashicorp/go-hclog v1.2.0 // indirect
+	github.com/hashicorp/go-immutable-radix v1.3.1 // indirect
+	github.com/hashicorp/go-msgpack v1.1.5 // indirect
+	github.com/hashicorp/golang-lru v0.5.4 // indirect
 	github.com/jcmturner/aescts/v2 v2.0.0 // indirect
 	github.com/jcmturner/dnsutils/v2 v2.0.0 // indirect
 	github.com/jcmturner/goidentity/v6 v6.0.1 // indirect
-	github.com/jcmturner/rpc/v2 v2.0.2 // indirect
+	github.com/jcmturner/rpc/v2 v2.0.3 // indirect
+	github.com/josharian/intern v1.0.0 // indirect
+	github.com/klauspost/cpuid/v2 v2.0.6 // indirect
+	github.com/mattn/go-colorable v0.1.12 // indirect
 	github.com/mattn/go-runewidth v0.0.7 // indirect
 	github.com/mattn/go-sqlite3 v2.0.1+incompatible // indirect
+	github.com/nats-io/nats.go v1.13.1-0.20220121202836-972a071d373d // indirect
+	github.com/nats-io/nkeys v0.3.0 // indirect
+	github.com/nats-io/nuid v1.0.1 // indirect
+	github.com/pelletier/go-toml/v2 v2.0.0-beta.8 // indirect
 	github.com/philhofer/fwd v1.1.1 // indirect
+	github.com/rwcarlsen/goexif v0.0.0-20190401172101-9e8deecbddbd // indirect
 	github.com/spf13/pflag v1.0.5 // indirect
+	github.com/subosito/gotenv v1.2.0 // indirect
 	github.com/tinylib/msgp v1.1.6 // indirect
-	go.etcd.io/etcd/api/v3 v3.5.0 // indirect
-	go.etcd.io/etcd/client/pkg/v3 v3.5.0 // indirect
-	go.uber.org/atomic v1.7.0 // indirect
-	go.uber.org/multierr v1.7.0 // indirect
-	go.uber.org/zap v1.17.0 // indirect
-	golang.org/x/mod v0.4.2 // indirect
+	go.etcd.io/etcd/api/v3 v3.5.3 // indirect
+	go.etcd.io/etcd/client/pkg/v3 v3.5.3 // indirect
+	go.uber.org/atomic v1.9.0 // indirect
+	go.uber.org/multierr v1.8.0 // indirect
+	go.uber.org/zap v1.21.0 // indirect
+	golang.org/x/mod v0.6.0-dev.0.20211013180041-c96bc1413d57 // indirect
 	golang.org/x/sync v0.0.0-20210220032951-036812b2e83c // indirect
+	gopkg.in/ini.v1 v1.66.4 // indirect
 	gopkg.in/yaml.v2 v2.4.0 // indirect
 	gopkg.in/yaml.v3 v3.0.0-20210107192922-496545a6307b // indirect
+	lukechampine.com/uint128 v1.1.1 // indirect
 )
 
-// replace github.com/seaweedfs/fuse => /Users/chrislu/go/src/github.com/seaweedfs/fuse
 // replace github.com/chrislusf/raft => /Users/chrislu/go/src/github.com/chrislusf/raft

k8s/helm_charts2/Chart.yaml

@@ -1,5 +1,5 @@
 apiVersion: v1
 description: SeaweedFS
 name: seaweedfs
-appVersion: "2.88"
-version: "2.88"
+appVersion: "2.99"
+version: "2.99"

k8s/helm_charts2/templates/s3-deployment.yaml

@@ -130,7 +130,7 @@ spec:
               name: swfs-s3
           readinessProbe:
             httpGet:
-              path: /
+              path: /status
               port: {{ .Values.s3.port }}
               scheme: HTTP
             initialDelaySeconds: 15
@@ -140,7 +140,7 @@ spec:
             timeoutSeconds: 10
           livenessProbe:
             httpGet:
-              path: /
+              path: /status
               port: {{ .Values.s3.port }}
               scheme: HTTP
             initialDelaySeconds: 20

k8s/helm_charts2/values.yaml

@@ -41,8 +41,7 @@ master:
   grpcPort: 19333
   ipBind: "0.0.0.0"
   volumePreallocate: false
-  #Master stops directing writes to oversized volumes
-  volumeSizeLimitMB: 30000
+  volumeSizeLimitMB: 1000
   loggingOverrideLevel: null
   #number of seconds between heartbeats, default 5
   pulseSeconds: null

other/java/client/src/main/proto/filer.proto

@@ -48,6 +48,9 @@ service SeaweedFiler {
     rpc Statistics (StatisticsRequest) returns (StatisticsResponse) {
     }
 
+    rpc Ping (PingRequest) returns (PingResponse) {
+    }
+
     rpc GetFilerConfiguration (GetFilerConfigurationRequest) returns (GetFilerConfigurationResponse) {
     }
 
@@ -171,6 +174,8 @@ message FuseAttributes {
     string symlink_target = 13;
     bytes md5 = 14;
     string disk_type = 15;
+    uint32 rdev = 16;
+    uint64 inode = 17;
 }
 
 message CreateEntryRequest {
@@ -179,6 +184,7 @@ message CreateEntryRequest {
     bool o_excl = 3;
     bool is_from_other_cluster = 4;
     repeated int32 signatures = 5;
+    bool skip_check_parent_directory = 6;
 }
 
 message CreateEntryResponse {
@@ -308,6 +314,16 @@ message StatisticsResponse {
     uint64 file_count = 6;
 }
 
+message PingRequest {
+    string target = 1; // default to ping itself
+    string target_type = 2;
+}
+message PingResponse {
+    int64 start_time_ns = 1;
+    int64 remote_time_ns = 2;
+    int64 stop_time_ns = 3;
+}
+
 message GetFilerConfigurationRequest {
 }
 message GetFilerConfigurationResponse {

other/java/hdfs-over-ftp/pom.xml

@@ -36,7 +36,7 @@
         <dependency>
             <groupId>org.apache.hadoop</groupId>
             <artifactId>hadoop-common</artifactId>
-            <version>3.2.2</version>
+            <version>3.2.3</version>
         </dependency>
         <dependency>
             <groupId>org.apache.hadoop</groupId>

other/java/hdfs3/pom.xml

@@ -6,7 +6,7 @@
 
     <properties>
         <seaweedfs.client.version>2.85</seaweedfs.client.version>
-        <hadoop.version>3.1.4</hadoop.version>
+        <hadoop.version>3.2.3</hadoop.version>
     </properties>
 
     <groupId>com.github.chrislusf</groupId>

weed/Makefile

@@ -21,10 +21,6 @@ debug_mount:
 	go build -gcflags="all=-N -l"
 	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- -v=4 mount -dir=~/tmp/mm -cacheCapacityMB=0 -filer.path=/ -umask=000
 
-debug_mount2:
-	go build -gcflags="all=-N -l"
-	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- -v=4 mount2 -dir=~/tmp/mm -cacheCapacityMB=0 -filer.path=/ -umask=000
-
 debug_server:
 	go build -gcflags="all=-N -l"
 	dlv --listen=:2345 --headless=true --api-version=2 --accept-multiclient exec weed -- server -dir=~/tmp/99 -filer -volume.port=8343 -s3 -volume.max=0 -master.volumeSizeLimitMB=1024 -volume.preStopSeconds=1

weed/cluster/cluster.go

@@ -9,9 +9,10 @@ import (
 )
 
 const (
-	MasterType = "master"
-	FilerType  = "filer"
-	BrokerType = "broker"
+	MasterType       = "master"
+	VolumeServerType = "volumeServer"
+	FilerType        = "filer"
+	BrokerType       = "broker"
 )
 
 type ClusterNode struct {
@@ -80,6 +81,15 @@ func (cluster *Cluster) AddClusterNode(nodeType string, address pb.ServerAddress
 			},
 		}
 	case MasterType:
+		return []*master_pb.KeepConnectedResponse{
+			{
+				ClusterNodeUpdate: &master_pb.ClusterNodeUpdate{
+					NodeType: nodeType,
+					Address:  string(address),
+					IsAdd:    true,
+				},
+			},
+		}
 	}
 	return nil
 }
@@ -119,6 +129,15 @@ func (cluster *Cluster) RemoveClusterNode(nodeType string, address pb.ServerAddr
 			}
 		}
 	case MasterType:
+		return []*master_pb.KeepConnectedResponse{
+			{
+				ClusterNodeUpdate: &master_pb.ClusterNodeUpdate{
+					NodeType: nodeType,
+					Address:  string(address),
+					IsAdd:    false,
+				},
+			},
+		}
 	}
 	return nil
 }

weed/command/benchmark.go

@@ -129,7 +129,7 @@ func runBenchmark(cmd *Command, args []string) bool {
 		defer pprof.StopCPUProfile()
 	}
 
-	b.masterClient = wdclient.NewMasterClient(b.grpcDialOption, "client", "", "", pb.ServerAddresses(*b.masters).ToAddresses())
+	b.masterClient = wdclient.NewMasterClient(b.grpcDialOption, "client", "", "", pb.ServerAddresses(*b.masters).ToAddressMap())
 	go b.masterClient.KeepConnectedToMaster()
 	b.masterClient.WaitUntilConnected()
 

weed/command/command.go

@@ -30,7 +30,6 @@ var Commands = []*Command{
 	cmdMaster,
 	cmdMasterFollower,
 	cmdMount,
-	cmdMount2,
 	cmdS3,
 	cmdIam,
 	cmdMsgBroker,

weed/command/filer.go

@@ -2,8 +2,10 @@ package command
 
 import (
 	"fmt"
+	"net"
 	"net/http"
 	"os"
+	"runtime"
 	"time"
 
 	"google.golang.org/grpc/reflection"
@@ -12,7 +14,7 @@ import (
 	"github.com/chrislusf/seaweedfs/weed/pb"
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
 	"github.com/chrislusf/seaweedfs/weed/security"
-	"github.com/chrislusf/seaweedfs/weed/server"
+	weed_server "github.com/chrislusf/seaweedfs/weed/server"
 	stats_collect "github.com/chrislusf/seaweedfs/weed/stats"
 	"github.com/chrislusf/seaweedfs/weed/util"
 )
@@ -28,7 +30,7 @@ var (
 )
 
 type FilerOptions struct {
-	masters                 []pb.ServerAddress
+	masters                 map[string]pb.ServerAddress
 	mastersString           *string
 	ip                      *string
 	bindIp                  *string
@@ -51,6 +53,7 @@ type FilerOptions struct {
 	concurrentUploadLimitMB *int
 	debug                   *bool
 	debugPort               *int
+	localSocket             *string
 }
 
 func init() {
@@ -58,7 +61,7 @@ func init() {
 	f.mastersString = cmdFiler.Flag.String("master", "localhost:9333", "comma-separated master servers")
 	f.collection = cmdFiler.Flag.String("collection", "", "all data will be stored in this default collection")
 	f.ip = cmdFiler.Flag.String("ip", util.DetectedHostAddress(), "filer server http listen ip address")
-	f.bindIp = cmdFiler.Flag.String("ip.bind", "", "ip address to bind to")
+	f.bindIp = cmdFiler.Flag.String("ip.bind", "", "ip address to bind to. If empty, default to same as -ip option.")
 	f.port = cmdFiler.Flag.Int("port", 8888, "filer server http listen port")
 	f.portGrpc = cmdFiler.Flag.Int("port.grpc", 0, "filer server grpc listen port")
 	f.publicPort = cmdFiler.Flag.Int("port.readonly", 0, "readonly port opened to public")
@@ -76,6 +79,7 @@ func init() {
 	f.concurrentUploadLimitMB = cmdFiler.Flag.Int("concurrentUploadLimitMB", 128, "limit total concurrent upload size")
 	f.debug = cmdFiler.Flag.Bool("debug", false, "serves runtime profiling data, e.g., http://localhost:<debug.port>/debug/pprof/goroutine?debug=2")
 	f.debugPort = cmdFiler.Flag.Int("debug.port", 6060, "http port for debugging")
+	f.localSocket = cmdFiler.Flag.String("localSocket", "", "default to /tmp/seaweedfs-filer-<port>.sock")
 
 	// start s3 on filer
 	filerStartS3 = cmdFiler.Flag.Bool("s3", false, "whether to start S3 gateway")
@@ -86,6 +90,7 @@ func init() {
 	filerS3Options.config = cmdFiler.Flag.String("s3.config", "", "path to the config file")
 	filerS3Options.auditLogConfig = cmdFiler.Flag.String("s3.auditLogConfig", "", "path to the audit log config file")
 	filerS3Options.allowEmptyFolder = cmdFiler.Flag.Bool("s3.allowEmptyFolder", true, "allow empty folders")
+	filerS3Options.allowDeleteBucketNotEmpty = cmdFiler.Flag.Bool("s3.allowDeleteBucketNotEmpty", true, "allow recursive deleting all entries along with bucket")
 
 	// start webdav on filer
 	filerStartWebDav = cmdFiler.Flag.Bool("webdav", false, "whether to start webdav gateway")
@@ -96,10 +101,11 @@ func init() {
 	filerWebDavOptions.tlsPrivateKey = cmdFiler.Flag.String("webdav.key.file", "", "path to the TLS private key file")
 	filerWebDavOptions.tlsCertificate = cmdFiler.Flag.String("webdav.cert.file", "", "path to the TLS certificate file")
 	filerWebDavOptions.cacheDir = cmdFiler.Flag.String("webdav.cacheDir", os.TempDir(), "local cache directory for file chunks")
-	filerWebDavOptions.cacheSizeMB = cmdFiler.Flag.Int64("webdav.cacheCapacityMB", 1000, "local cache capacity in MB")
+	filerWebDavOptions.cacheSizeMB = cmdFiler.Flag.Int64("webdav.cacheCapacityMB", 0, "local cache capacity in MB")
 
 	// start iam on filer
 	filerStartIam = cmdFiler.Flag.Bool("iam", false, "whether to start IAM service")
+	filerIamOptions.ip = cmdFiler.Flag.String("iam.ip", *f.ip, "iam server http listen ip address")
 	filerIamOptions.port = cmdFiler.Flag.Int("iam.port", 8111, "iam server http listen port")
 }
 
@@ -139,11 +145,14 @@ func runFiler(cmd *Command, args []string) bool {
 	if *filerStartS3 {
 		filerS3Options.filer = &filerAddress
 		filerS3Options.bindIp = f.bindIp
+		filerS3Options.localFilerSocket = f.localSocket
 		go func() {
 			time.Sleep(startDelay * time.Second)
 			filerS3Options.startS3Server()
 		}()
 		startDelay++
+	} else {
+		*f.localSocket = ""
 	}
 
 	if *filerStartWebDav {
@@ -164,7 +173,7 @@ func runFiler(cmd *Command, args []string) bool {
 		}()
 	}
 
-	f.masters = pb.ServerAddresses(*f.mastersString).ToAddresses()
+	f.masters = pb.ServerAddresses(*f.mastersString).ToAddressMap()
 
 	f.startFiler()
 
@@ -182,6 +191,9 @@ func (fo *FilerOptions) startFiler() {
 	if *fo.portGrpc == 0 {
 		*fo.portGrpc = 10000 + *fo.port
 	}
+	if *fo.bindIp == "" {
+		*fo.bindIp = *fo.ip
+	}
 
 	defaultLevelDbDirectory := util.ResolvePath(*fo.defaultLevelDbDirectory + "/filerldb2")
 
@@ -210,7 +222,7 @@ func (fo *FilerOptions) startFiler() {
 	if *fo.publicPort != 0 {
 		publicListeningAddress := util.JoinHostPort(*fo.bindIp, *fo.publicPort)
 		glog.V(0).Infoln("Start Seaweed filer server", util.Version(), "public at", publicListeningAddress)
-		publicListener, e := util.NewListener(publicListeningAddress, 0)
+		publicListener, localPublicListner, e := util.NewIpAndLocalListeners(*fo.bindIp, *fo.publicPort, 0)
 		if e != nil {
 			glog.Fatalf("Filer server public listener error on port %d:%v", *fo.publicPort, e)
 		}
@@ -219,11 +231,18 @@ func (fo *FilerOptions) startFiler() {
 				glog.Fatalf("Volume server fail to serve public: %v", e)
 			}
 		}()
+		if localPublicListner != nil {
+			go func() {
+				if e := http.Serve(localPublicListner, publicVolumeMux); e != nil {
+					glog.Errorf("Volume server fail to serve public: %v", e)
+				}
+			}()
+		}
 	}
 
 	glog.V(0).Infof("Start Seaweed Filer %s at %s:%d", util.Version(), *fo.ip, *fo.port)
-	filerListener, e := util.NewListener(
-		util.JoinHostPort(*fo.bindIp, *fo.port),
+	filerListener, filerLocalListener, e := util.NewIpAndLocalListeners(
+		*fo.bindIp, *fo.port,
 		time.Duration(10)*time.Second,
 	)
 	if e != nil {
@@ -232,16 +251,42 @@ func (fo *FilerOptions) startFiler() {
 
 	// starting grpc server
 	grpcPort := *fo.portGrpc
-	grpcL, err := util.NewListener(util.JoinHostPort(*fo.bindIp, grpcPort), 0)
+	grpcL, grpcLocalL, err := util.NewIpAndLocalListeners(*fo.bindIp, grpcPort, 0)
 	if err != nil {
 		glog.Fatalf("failed to listen on grpc port %d: %v", grpcPort, err)
 	}
 	grpcS := pb.NewGrpcServer(security.LoadServerTLS(util.GetViper(), "grpc.filer"))
 	filer_pb.RegisterSeaweedFilerServer(grpcS, fs)
 	reflection.Register(grpcS)
+	if grpcLocalL != nil {
+		go grpcS.Serve(grpcLocalL)
+	}
 	go grpcS.Serve(grpcL)
 
 	httpS := &http.Server{Handler: defaultMux}
+	if runtime.GOOS != "windows" {
+		if *fo.localSocket == "" {
+			*fo.localSocket = fmt.Sprintf("/tmp/seaweefs-filer-%d.sock", *fo.port)
+			if err := os.Remove(*fo.localSocket); err != nil && !os.IsNotExist(err) {
+				glog.Fatalf("Failed to remove %s, error: %s", *fo.localSocket, err.Error())
+			}
+		}
+		go func() {
+			// start on local unix socket
+			filerSocketListener, err := net.Listen("unix", *fo.localSocket)
+			if err != nil {
+				glog.Fatalf("Failed to listen on %s: %v", *fo.localSocket, err)
+			}
+			httpS.Serve(filerSocketListener)
+		}()
+	}
+	if filerLocalListener != nil {
+		go func() {
+			if err := httpS.Serve(filerLocalListener); err != nil {
+				glog.Errorf("Filer Fail to serve: %v", e)
+			}
+		}()
+	}
 	if err := httpS.Serve(filerListener); err != nil {
 		glog.Fatalf("Filer Fail to serve: %v", e)
 	}

weed/command/filer_meta_backup.go

@@ -162,24 +162,21 @@ func (metaBackup *FilerMetaBackupOptions) streamMetadataBackup() error {
 		ctx := context.Background()
 		message := resp.EventNotification
 
-		if message.OldEntry == nil && message.NewEntry == nil {
+		if filer_pb.IsEmpty(resp) {
 			return nil
-		}
-		if message.OldEntry == nil && message.NewEntry != nil {
+		} else if filer_pb.IsCreate(resp) {
 			println("+", util.FullPath(message.NewParentPath).Child(message.NewEntry.Name))
 			entry := filer.FromPbEntry(message.NewParentPath, message.NewEntry)
 			return store.InsertEntry(ctx, entry)
-		}
-		if message.OldEntry != nil && message.NewEntry == nil {
+		} else if filer_pb.IsDelete(resp) {
 			println("-", util.FullPath(resp.Directory).Child(message.OldEntry.Name))
 			return store.DeleteEntry(ctx, util.FullPath(resp.Directory).Child(message.OldEntry.Name))
-		}
-		if message.OldEntry != nil && message.NewEntry != nil {
-			if resp.Directory == message.NewParentPath && message.OldEntry.Name == message.NewEntry.Name {
-				println("~", util.FullPath(message.NewParentPath).Child(message.NewEntry.Name))
-				entry := filer.FromPbEntry(message.NewParentPath, message.NewEntry)
-				return store.UpdateEntry(ctx, entry)
-			}
+		} else if filer_pb.IsUpdate(resp) {
+			println("~", util.FullPath(message.NewParentPath).Child(message.NewEntry.Name))
+			entry := filer.FromPbEntry(message.NewParentPath, message.NewEntry)
+			return store.UpdateEntry(ctx, entry)
+		} else {
+			// renaming
 			println("-", util.FullPath(resp.Directory).Child(message.OldEntry.Name))
 			if err := store.DeleteEntry(ctx, util.FullPath(resp.Directory).Child(message.OldEntry.Name)); err != nil {
 				return err

weed/command/filer_meta_tail.go

@@ -74,7 +74,7 @@ func runFilerMetaTail(cmd *Command, args []string) bool {
 	}
 
 	shouldPrint := func(resp *filer_pb.SubscribeMetadataResponse) bool {
-		if resp.EventNotification.OldEntry == nil && resp.EventNotification.NewEntry == nil {
+		if filer_pb.IsEmpty(resp) {
 			return false
 		}
 		if filterFunc == nil {

weed/command/filer_remote_gateway_buckets.go

@@ -174,10 +174,10 @@ func (option *RemoteGatewayOptions) makeBucketedEventProcessor(filerSource *sour
 			return handleEtcRemoteChanges(resp)
 		}
 
-		if message.OldEntry == nil && message.NewEntry == nil {
+		if filer_pb.IsEmpty(resp) {
 			return nil
 		}
-		if message.OldEntry == nil && message.NewEntry != nil {
+		if filer_pb.IsCreate(resp) {
 			if message.NewParentPath == option.bucketsDir {
 				return handleCreateBucket(message.NewEntry)
 			}
@@ -212,7 +212,7 @@ func (option *RemoteGatewayOptions) makeBucketedEventProcessor(filerSource *sour
 			}
 			return updateLocalEntry(&remoteSyncOptions, message.NewParentPath, message.NewEntry, remoteEntry)
 		}
-		if message.OldEntry != nil && message.NewEntry == nil {
+		if filer_pb.IsDelete(resp) {
 			if resp.Directory == option.bucketsDir {
 				return handleDeleteBucket(message.OldEntry)
 			}

weed/command/filer_remote_sync_dir.go

@@ -91,10 +91,10 @@ func makeEventProcessor(remoteStorage *remote_pb.RemoteConf, mountedDir string,
 			return handleEtcRemoteChanges(resp)
 		}
 
-		if message.OldEntry == nil && message.NewEntry == nil {
+		if filer_pb.IsEmpty(resp) {
 			return nil
 		}
-		if message.OldEntry == nil && message.NewEntry != nil {
+		if filer_pb.IsCreate(resp) {
 			if !filer.HasData(message.NewEntry) {
 				return nil
 			}
@@ -115,7 +115,7 @@ func makeEventProcessor(remoteStorage *remote_pb.RemoteConf, mountedDir string,
 			}
 			return updateLocalEntry(&remoteSyncOptions, message.NewParentPath, message.NewEntry, remoteEntry)
 		}
-		if message.OldEntry != nil && message.NewEntry == nil {
+		if filer_pb.IsDelete(resp) {
 			glog.V(2).Infof("delete: %+v", resp)
 			dest := toRemoteStorageLocation(util.FullPath(mountedDir), util.NewFullPath(resp.Directory, message.OldEntry.Name), remoteStorageMountLocation)
 			if message.OldEntry.IsDirectory {

weed/command/filer_sync.go

@@ -262,16 +262,19 @@ func genProcessFunction(sourcePath string, targetPath string, dataSink sink.Repl
 		}
 
 		// handle deletions
-		if message.OldEntry != nil && message.NewEntry == nil {
+		if filer_pb.IsDelete(resp) {
 			if !strings.HasPrefix(string(sourceOldKey), sourcePath) {
 				return nil
 			}
 			key := buildKey(dataSink, message, targetPath, sourceOldKey, sourcePath)
-			return dataSink.DeleteEntry(key, message.OldEntry.IsDirectory, message.DeleteChunks, message.Signatures)
+			if !dataSink.IsIncremental() {
+				return dataSink.DeleteEntry(key, message.OldEntry.IsDirectory, message.DeleteChunks, message.Signatures)
+			}
+			return nil
 		}
 
 		// handle new entries
-		if message.OldEntry == nil && message.NewEntry != nil {
+		if filer_pb.IsCreate(resp) {
 			if !strings.HasPrefix(string(sourceNewKey), sourcePath) {
 				return nil
 			}
@@ -280,7 +283,7 @@ func genProcessFunction(sourcePath string, targetPath string, dataSink sink.Repl
 		}
 
 		// this is something special?
-		if message.OldEntry == nil && message.NewEntry == nil {
+		if filer_pb.IsEmpty(resp) {
 			return nil
 		}
 

weed/command/iam.go

@@ -22,6 +22,7 @@ var (
 type IamOptions struct {
 	filer   *string
 	masters *string
+	ip      *string
 	port    *int
 }
 
@@ -29,6 +30,7 @@ func init() {
 	cmdIam.Run = runIam // break init cycle
 	iamStandaloneOptions.filer = cmdIam.Flag.String("filer", "localhost:8888", "filer server address")
 	iamStandaloneOptions.masters = cmdIam.Flag.String("master", "localhost:9333", "comma-separated master servers")
+	iamStandaloneOptions.ip = cmdIam.Flag.String("ip", util.DetectedHostAddress(), "iam server http listen ip address")
 	iamStandaloneOptions.port = cmdIam.Flag.Int("port", 8111, "iam server http listen port")
 }
 
@@ -65,7 +67,7 @@ func (iamopt *IamOptions) startIamServer() bool {
 		}
 	}
 
-	masters := pb.ServerAddresses(*iamopt.masters).ToAddresses()
+	masters := pb.ServerAddresses(*iamopt.masters).ToAddressMap()
 	router := mux.NewRouter().SkipClean(true)
 	_, iamApiServer_err := iamapi.NewIamApiServer(router, &iamapi.IamServerOption{
 		Masters:        masters,
@@ -81,12 +83,19 @@ func (iamopt *IamOptions) startIamServer() bool {
 	httpS := &http.Server{Handler: router}
 
 	listenAddress := fmt.Sprintf(":%d", *iamopt.port)
-	iamApiListener, err := util.NewListener(listenAddress, time.Duration(10)*time.Second)
+	iamApiListener, iamApiLocalListener, err := util.NewIpAndLocalListeners(*iamopt.ip, *iamopt.port, time.Duration(10)*time.Second)
 	if err != nil {
 		glog.Fatalf("IAM API Server listener on %s error: %v", listenAddress, err)
 	}
 
 	glog.V(0).Infof("Start Seaweed IAM API Server %s at http port %d", util.Version(), *iamopt.port)
+	if iamApiLocalListener != nil {
+		go func() {
+			if err = httpS.Serve(iamApiLocalListener); err != nil {
+				glog.Errorf("IAM API Server Fail to serve: %v", err)
+			}
+		}()
+	}
 	if err = httpS.Serve(iamApiListener); err != nil {
 		glog.Fatalf("IAM API Server Fail to serve: %v", err)
 	}

weed/command/master.go

@@ -1,15 +1,17 @@
 package command
 
 import (
+	"golang.org/x/exp/slices"
+	"net/http"
+	"os"
+	"strings"
+	"time"
+
 	"github.com/chrislusf/raft/protobuf"
 	stats_collect "github.com/chrislusf/seaweedfs/weed/stats"
 	"github.com/gorilla/mux"
+	"github.com/spf13/viper"
 	"google.golang.org/grpc/reflection"
-	"net/http"
-	"os"
-	"sort"
-	"strings"
-	"time"
 
 	"github.com/chrislusf/seaweedfs/weed/util/grace"
 
@@ -17,7 +19,7 @@ import (
 	"github.com/chrislusf/seaweedfs/weed/pb"
 	"github.com/chrislusf/seaweedfs/weed/pb/master_pb"
 	"github.com/chrislusf/seaweedfs/weed/security"
-	"github.com/chrislusf/seaweedfs/weed/server"
+	weed_server "github.com/chrislusf/seaweedfs/weed/server"
 	"github.com/chrislusf/seaweedfs/weed/storage/backend"
 	"github.com/chrislusf/seaweedfs/weed/util"
 )
@@ -44,6 +46,10 @@ type MasterOptions struct {
 	metricsIntervalSec *int
 	raftResumeState    *bool
 	metricsHttpPort    *int
+	heartbeatInterval  *time.Duration
+	electionTimeout    *time.Duration
+	raftHashicorp      *bool
+	raftBootstrap      *bool
 }
 
 func init() {
@@ -51,7 +57,7 @@ func init() {
 	m.port = cmdMaster.Flag.Int("port", 9333, "http listen port")
 	m.portGrpc = cmdMaster.Flag.Int("port.grpc", 0, "grpc listen port")
 	m.ip = cmdMaster.Flag.String("ip", util.DetectedHostAddress(), "master <ip>|<server> address, also used as identifier")
-	m.ipBind = cmdMaster.Flag.String("ip.bind", "", "ip address to bind to")
+	m.ipBind = cmdMaster.Flag.String("ip.bind", "", "ip address to bind to. If empty, default to same as -ip option.")
 	m.metaFolder = cmdMaster.Flag.String("mdir", os.TempDir(), "data directory to store meta data")
 	m.peers = cmdMaster.Flag.String("peers", "", "all master nodes in comma separated ip:port list, example: 127.0.0.1:9093,127.0.0.1:9094,127.0.0.1:9095")
 	m.volumeSizeLimitMB = cmdMaster.Flag.Uint("volumeSizeLimitMB", 30*1000, "Master stops directing writes to oversized volumes.")
@@ -65,6 +71,10 @@ func init() {
 	m.metricsIntervalSec = cmdMaster.Flag.Int("metrics.intervalSeconds", 15, "Prometheus push interval in seconds")
 	m.metricsHttpPort = cmdMaster.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
 	m.raftResumeState = cmdMaster.Flag.Bool("resumeState", false, "resume previous state on start master server")
+	m.heartbeatInterval = cmdMaster.Flag.Duration("heartbeatInterval", 300*time.Millisecond, "heartbeat interval of master servers, and will be randomly multiplied by [1, 1.25)")
+	m.electionTimeout = cmdMaster.Flag.Duration("electionTimeout", 10*time.Second, "election timeout of master servers")
+	m.raftHashicorp = cmdMaster.Flag.Bool("raftHashicorp", false, "use hashicorp raft")
+	m.raftBootstrap = cmdMaster.Flag.Bool("raftBootstrap", false, "Whether to bootstrap the Raft cluster")
 }
 
 var cmdMaster = &Command{
@@ -120,52 +130,123 @@ func startMaster(masterOption MasterOptions, masterWhiteList []string) {
 	if *masterOption.portGrpc == 0 {
 		*masterOption.portGrpc = 10000 + *masterOption.port
 	}
+	if *masterOption.ipBind == "" {
+		*masterOption.ipBind = *masterOption.ip
+	}
 
 	myMasterAddress, peers := checkPeers(*masterOption.ip, *masterOption.port, *masterOption.portGrpc, *masterOption.peers)
 
+	masterPeers := make(map[string]pb.ServerAddress)
+	for _, peer := range peers {
+		masterPeers[string(peer)] = peer
+	}
+
 	r := mux.NewRouter()
-	ms := weed_server.NewMasterServer(r, masterOption.toMasterOption(masterWhiteList), peers)
+	ms := weed_server.NewMasterServer(r, masterOption.toMasterOption(masterWhiteList), masterPeers)
 	listeningAddress := util.JoinHostPort(*masterOption.ipBind, *masterOption.port)
 	glog.V(0).Infof("Start Seaweed Master %s at %s", util.Version(), listeningAddress)
-	masterListener, e := util.NewListener(listeningAddress, 0)
+	masterListener, masterLocalListner, e := util.NewIpAndLocalListeners(*masterOption.ipBind, *masterOption.port, 0)
 	if e != nil {
 		glog.Fatalf("Master startup error: %v", e)
 	}
+
 	// start raftServer
-	raftServer, err := weed_server.NewRaftServer(security.LoadClientTLS(util.GetViper(), "grpc.master"),
-		peers, myMasterAddress, util.ResolvePath(*masterOption.metaFolder), ms.Topo, *masterOption.raftResumeState)
-	if raftServer == nil {
-		glog.Fatalf("please verify %s is writable, see https://github.com/chrislusf/seaweedfs/issues/717: %s", *masterOption.metaFolder, err)
+	raftServerOption := &weed_server.RaftServerOption{
+		GrpcDialOption:    security.LoadClientTLS(util.GetViper(), "grpc.master"),
+		Peers:             masterPeers,
+		ServerAddr:        myMasterAddress,
+		DataDir:           util.ResolvePath(*masterOption.metaFolder),
+		Topo:              ms.Topo,
+		RaftResumeState:   *masterOption.raftResumeState,
+		HeartbeatInterval: *masterOption.heartbeatInterval,
+		ElectionTimeout:   *masterOption.electionTimeout,
+		RaftBootstrap:     *m.raftBootstrap,
+	}
+	var raftServer *weed_server.RaftServer
+	var err error
+	if *m.raftHashicorp {
+		if raftServer, err = weed_server.NewHashicorpRaftServer(raftServerOption); err != nil {
+			glog.Fatalf("NewHashicorpRaftServer: %s", err)
+		}
+	} else {
+		raftServer, err = weed_server.NewRaftServer(raftServerOption)
+		if raftServer == nil {
+			glog.Fatalf("please verify %s is writable, see https://github.com/chrislusf/seaweedfs/issues/717: %s", *masterOption.metaFolder, err)
+		}
 	}
 	ms.SetRaftServer(raftServer)
 	r.HandleFunc("/cluster/status", raftServer.StatusHandler).Methods("GET")
+	if *m.raftHashicorp {
+		r.HandleFunc("/raft/stats", raftServer.StatsRaftHandler).Methods("GET")
+	}
 	// starting grpc server
 	grpcPort := *masterOption.portGrpc
-	grpcL, err := util.NewListener(util.JoinHostPort(*masterOption.ipBind, grpcPort), 0)
+	grpcL, grpcLocalL, err := util.NewIpAndLocalListeners(*masterOption.ipBind, grpcPort, 0)
 	if err != nil {
 		glog.Fatalf("master failed to listen on grpc port %d: %v", grpcPort, err)
 	}
 	grpcS := pb.NewGrpcServer(security.LoadServerTLS(util.GetViper(), "grpc.master"))
 	master_pb.RegisterSeaweedServer(grpcS, ms)
-	protobuf.RegisterRaftServer(grpcS, raftServer)
+	if *m.raftHashicorp {
+		raftServer.TransportManager.Register(grpcS)
+	} else {
+		protobuf.RegisterRaftServer(grpcS, raftServer)
+	}
 	reflection.Register(grpcS)
 	glog.V(0).Infof("Start Seaweed Master %s grpc server at %s:%d", util.Version(), *masterOption.ipBind, grpcPort)
+	if grpcLocalL != nil {
+		go grpcS.Serve(grpcLocalL)
+	}
 	go grpcS.Serve(grpcL)
 
-	go func() {
-		time.Sleep(1500 * time.Millisecond)
-		if ms.Topo.RaftServer.Leader() == "" && ms.Topo.RaftServer.IsLogEmpty() && isTheFirstOne(myMasterAddress, peers) {
-			if ms.MasterClient.FindLeaderFromOtherPeers(myMasterAddress) == "" {
-				raftServer.DoJoinCommand()
+	timeSleep := 1500 * time.Millisecond
+	if !*m.raftHashicorp {
+		go func() {
+			time.Sleep(timeSleep)
+			if ms.Topo.RaftServer.Leader() == "" && ms.Topo.RaftServer.IsLogEmpty() && isTheFirstOne(myMasterAddress, peers) {
+				if ms.MasterClient.FindLeaderFromOtherPeers(myMasterAddress) == "" {
+					raftServer.DoJoinCommand()
+				}
 			}
-		}
-	}()
+		}()
+	}
 
 	go ms.MasterClient.KeepConnectedToMaster()
 
 	// start http server
+	var (
+		clientCertFile,
+		certFile,
+		keyFile string
+	)
+	useTLS := false
+	useMTLS := false
+
+	if viper.GetString("https.master.key") != "" {
+		useTLS = true
+		certFile = viper.GetString("https.master.cert")
+		keyFile = viper.GetString("https.master.key")
+	}
+
+	if viper.GetString("https.master.ca") != "" {
+		useMTLS = true
+		clientCertFile = viper.GetString("https.master.ca")
+	}
+
 	httpS := &http.Server{Handler: r}
-	go httpS.Serve(masterListener)
+	if masterLocalListner != nil {
+		go httpS.Serve(masterLocalListner)
+	}
+
+	if useMTLS {
+		httpS.TLSConfig = security.LoadClientTLSHTTP(clientCertFile)
+	}
+
+	if useTLS {
+		go httpS.ServeTLS(masterListener, certFile, keyFile)
+	} else {
+		go httpS.Serve(masterListener)
+	}
 
 	select {}
 }
@@ -193,8 +274,8 @@ func checkPeers(masterIp string, masterPort int, masterGrpcPort int, peers strin
 }
 
 func isTheFirstOne(self pb.ServerAddress, peers []pb.ServerAddress) bool {
-	sort.Slice(peers, func(i, j int) bool {
-		return strings.Compare(string(peers[i]), string(peers[j])) < 0
+	slices.SortFunc(peers, func(a, b pb.ServerAddress) bool {
+		return strings.Compare(string(a), string(b)) < 0
 	})
 	if len(peers) <= 0 {
 		return true

weed/command/master_follower.go

@@ -3,17 +3,18 @@ package command
 import (
 	"context"
 	"fmt"
+	"net/http"
+	"time"
+
 	"github.com/aws/aws-sdk-go/aws"
 	"github.com/chrislusf/seaweedfs/weed/glog"
 	"github.com/chrislusf/seaweedfs/weed/pb"
 	"github.com/chrislusf/seaweedfs/weed/pb/master_pb"
 	"github.com/chrislusf/seaweedfs/weed/security"
-	"github.com/chrislusf/seaweedfs/weed/server"
+	weed_server "github.com/chrislusf/seaweedfs/weed/server"
 	"github.com/chrislusf/seaweedfs/weed/util"
 	"github.com/gorilla/mux"
 	"google.golang.org/grpc/reflection"
-	"net/http"
-	"time"
 )
 
 var (
@@ -24,7 +25,7 @@ func init() {
 	cmdMasterFollower.Run = runMasterFollower // break init cycle
 	mf.port = cmdMasterFollower.Flag.Int("port", 9334, "http listen port")
 	mf.portGrpc = cmdMasterFollower.Flag.Int("port.grpc", 0, "grpc listen port")
-	mf.ipBind = cmdMasterFollower.Flag.String("ip.bind", "", "ip address to bind to")
+	mf.ipBind = cmdMasterFollower.Flag.String("ip.bind", "", "ip address to bind to. Default to localhost.")
 	mf.peers = cmdMasterFollower.Flag.String("masters", "localhost:9333", "all master nodes in comma separated ip:port list, example: 127.0.0.1:9093,127.0.0.1:9094,127.0.0.1:9095")
 
 	mf.ip = aws.String(util.DetectedHostAddress())
@@ -45,13 +46,13 @@ var cmdMasterFollower = &Command{
 	Short:     "start a master follower",
 	Long: `start a master follower to provide volume=>location mapping service
 
-	The master follower does not participate in master election. 
+	The master follower does not participate in master election.
 	It just follow the existing masters, and listen for any volume location changes.
 
 	In most cases, the master follower is not needed. In big data centers with thousands of volume
 	servers. In theory, the master may have trouble to keep up with the write requests and read requests.
 
-	The master follower can relieve the master from from read requests, which only needs to 
+	The master follower can relieve the master from from read requests, which only needs to
 	lookup a fileId or volumeId.
 
 	The master follower currently can handle fileId lookup requests:
@@ -82,7 +83,7 @@ func runMasterFollower(cmd *Command, args []string) bool {
 func startMasterFollower(masterOptions MasterOptions) {
 
 	// collect settings from main masters
-	masters := pb.ServerAddresses(*mf.peers).ToAddresses()
+	masters := pb.ServerAddresses(*mf.peers).ToAddressMap()
 
 	var err error
 	grpcDialOption := security.LoadClientTLS(util.GetViper(), "grpc.master")
@@ -111,18 +112,22 @@ func startMasterFollower(masterOptions MasterOptions) {
 	option := masterOptions.toMasterOption(nil)
 	option.IsFollower = true
 
+	if *masterOptions.ipBind == "" {
+		*masterOptions.ipBind = *masterOptions.ip
+	}
+
 	r := mux.NewRouter()
 	ms := weed_server.NewMasterServer(r, option, masters)
 	listeningAddress := util.JoinHostPort(*masterOptions.ipBind, *masterOptions.port)
 	glog.V(0).Infof("Start Seaweed Master %s at %s", util.Version(), listeningAddress)
-	masterListener, e := util.NewListener(listeningAddress, 0)
+	masterListener, masterLocalListner, e := util.NewIpAndLocalListeners(*masterOptions.ipBind, *masterOptions.port, 0)
 	if e != nil {
 		glog.Fatalf("Master startup error: %v", e)
 	}
 
 	// starting grpc server
 	grpcPort := *masterOptions.portGrpc
-	grpcL, err := util.NewListener(util.JoinHostPort(*masterOptions.ipBind, grpcPort), 0)
+	grpcL, grpcLocalL, err := util.NewIpAndLocalListeners(*masterOptions.ipBind, grpcPort, 0)
 	if err != nil {
 		glog.Fatalf("master failed to listen on grpc port %d: %v", grpcPort, err)
 	}
@@ -130,12 +135,18 @@ func startMasterFollower(masterOptions MasterOptions) {
 	master_pb.RegisterSeaweedServer(grpcS, ms)
 	reflection.Register(grpcS)
 	glog.V(0).Infof("Start Seaweed Master %s grpc server at %s:%d", util.Version(), *masterOptions.ip, grpcPort)
+	if grpcLocalL != nil {
+		go grpcS.Serve(grpcLocalL)
+	}
 	go grpcS.Serve(grpcL)
 
 	go ms.MasterClient.KeepConnectedToMaster()
 
 	// start http server
 	httpS := &http.Server{Handler: r}
+	if masterLocalListner != nil {
+		go httpS.Serve(masterLocalListner)
+	}
 	go httpS.Serve(masterListener)
 
 	select {}

weed/command/mount.go

@@ -11,6 +11,7 @@ type MountOptions struct {
 	dir                *string
 	dirAutoCreate      *bool
 	collection         *string
+	collectionQuota    *int
 	replication        *string
 	diskType           *string
 	ttlSec             *int
@@ -28,6 +29,7 @@ type MountOptions struct {
 	readOnly           *bool
 	debug              *bool
 	debugPort          *int
+	localSocket        *string
 }
 
 var (
@@ -44,13 +46,14 @@ func init() {
 	mountOptions.dir = cmdMount.Flag.String("dir", ".", "mount weed filer to this directory")
 	mountOptions.dirAutoCreate = cmdMount.Flag.Bool("dirAutoCreate", false, "auto create the directory to mount to")
 	mountOptions.collection = cmdMount.Flag.String("collection", "", "collection to create the files")
+	mountOptions.collectionQuota = cmdMount.Flag.Int("collectionQuotaMB", 0, "quota for the collection")
 	mountOptions.replication = cmdMount.Flag.String("replication", "", "replication(e.g. 000, 001) to create to files. If empty, let filer decide.")
 	mountOptions.diskType = cmdMount.Flag.String("disk", "", "[hdd|ssd|<tag>] hard drive or solid state drive or any tag")
 	mountOptions.ttlSec = cmdMount.Flag.Int("ttl", 0, "file ttl in seconds")
 	mountOptions.chunkSizeLimitMB = cmdMount.Flag.Int("chunkSizeLimitMB", 2, "local write buffer size, also chunk large files")
 	mountOptions.concurrentWriters = cmdMount.Flag.Int("concurrentWriters", 32, "limit concurrent goroutine writers if not 0")
 	mountOptions.cacheDir = cmdMount.Flag.String("cacheDir", os.TempDir(), "local cache directory for file chunks and meta data")
-	mountOptions.cacheSizeMB = cmdMount.Flag.Int64("cacheCapacityMB", 1000, "local file chunk cache capacity in MB (0 will disable cache)")
+	mountOptions.cacheSizeMB = cmdMount.Flag.Int64("cacheCapacityMB", 0, "local file chunk cache capacity in MB")
 	mountOptions.dataCenter = cmdMount.Flag.String("dataCenter", "", "prefer to write to the data center")
 	mountOptions.allowOthers = cmdMount.Flag.Bool("allowOthers", true, "allows other users to access the file system")
 	mountOptions.umaskString = cmdMount.Flag.String("umask", "022", "octal umask, e.g., 022, 0111")
@@ -61,6 +64,7 @@ func init() {
 	mountOptions.readOnly = cmdMount.Flag.Bool("readOnly", false, "read only")
 	mountOptions.debug = cmdMount.Flag.Bool("debug", false, "serves runtime profiling data, e.g., http://localhost:<debug.port>/debug/pprof/goroutine?debug=2")
 	mountOptions.debugPort = cmdMount.Flag.Int("debug.port", 6061, "http port for debugging")
+	mountOptions.localSocket = cmdMount.Flag.String("localSocket", "", "default to /tmp/seaweedfs-mount-<mount_dir_hash>.sock")
 
 	mountCpuProfile = cmdMount.Flag.String("cpuprofile", "", "cpu profile output file")
 	mountMemProfile = cmdMount.Flag.String("memprofile", "", "memory profile output file")

weed/command/mount2.go

@@ -1,83 +0,0 @@
-package command
-
-import (
-	"os"
-	"time"
-)
-
-type Mount2Options struct {
-	filer              *string
-	filerMountRootPath *string
-	dir                *string
-	dirAutoCreate      *bool
-	collection         *string
-	replication        *string
-	diskType           *string
-	ttlSec             *int
-	chunkSizeLimitMB   *int
-	concurrentWriters  *int
-	cacheDir           *string
-	cacheSizeMB        *int64
-	dataCenter         *string
-	allowOthers        *bool
-	umaskString        *string
-	nonempty           *bool
-	volumeServerAccess *string
-	uidMap             *string
-	gidMap             *string
-	readOnly           *bool
-	debug              *bool
-	debugPort          *int
-}
-
-var (
-	mount2Options Mount2Options
-)
-
-func init() {
-	cmdMount2.Run = runMount2 // break init cycle
-	mount2Options.filer = cmdMount2.Flag.String("filer", "localhost:8888", "comma-separated weed filer location")
-	mount2Options.filerMountRootPath = cmdMount2.Flag.String("filer.path", "/", "mount this remote path from filer server")
-	mount2Options.dir = cmdMount2.Flag.String("dir", ".", "mount weed filer to this directory")
-	mount2Options.dirAutoCreate = cmdMount2.Flag.Bool("dirAutoCreate", false, "auto create the directory to mount to")
-	mount2Options.collection = cmdMount2.Flag.String("collection", "", "collection to create the files")
-	mount2Options.replication = cmdMount2.Flag.String("replication", "", "replication(e.g. 000, 001) to create to files. If empty, let filer decide.")
-	mount2Options.diskType = cmdMount2.Flag.String("disk", "", "[hdd|ssd|<tag>] hard drive or solid state drive or any tag")
-	mount2Options.ttlSec = cmdMount2.Flag.Int("ttl", 0, "file ttl in seconds")
-	mount2Options.chunkSizeLimitMB = cmdMount2.Flag.Int("chunkSizeLimitMB", 2, "local write buffer size, also chunk large files")
-	mount2Options.concurrentWriters = cmdMount2.Flag.Int("concurrentWriters", 32, "limit concurrent goroutine writers if not 0")
-	mount2Options.cacheDir = cmdMount2.Flag.String("cacheDir", os.TempDir(), "local cache directory for file chunks and meta data")
-	mount2Options.cacheSizeMB = cmdMount2.Flag.Int64("cacheCapacityMB", 1000, "local file chunk cache capacity in MB (0 will disable cache)")
-	mount2Options.dataCenter = cmdMount2.Flag.String("dataCenter", "", "prefer to write to the data center")
-	mount2Options.allowOthers = cmdMount2.Flag.Bool("allowOthers", true, "allows other users to access the file system")
-	mount2Options.umaskString = cmdMount2.Flag.String("umask", "022", "octal umask, e.g., 022, 0111")
-	mount2Options.nonempty = cmdMount2.Flag.Bool("nonempty", false, "allows the mounting over a non-empty directory")
-	mount2Options.volumeServerAccess = cmdMount2.Flag.String("volumeServerAccess", "direct", "access volume servers by [direct|publicUrl|filerProxy]")
-	mount2Options.uidMap = cmdMount2.Flag.String("map.uid", "", "map local uid to uid on filer, comma-separated <local_uid>:<filer_uid>")
-	mount2Options.gidMap = cmdMount2.Flag.String("map.gid", "", "map local gid to gid on filer, comma-separated <local_gid>:<filer_gid>")
-	mount2Options.readOnly = cmdMount2.Flag.Bool("readOnly", false, "read only")
-	mount2Options.debug = cmdMount2.Flag.Bool("debug", false, "serves runtime profiling data, e.g., http://localhost:<debug.port>/debug/pprof/goroutine?debug=2")
-	mount2Options.debugPort = cmdMount2.Flag.Int("debug.port", 6061, "http port for debugging")
-
-	mountCpuProfile = cmdMount2.Flag.String("cpuprofile", "", "cpu profile output file")
-	mountMemProfile = cmdMount2.Flag.String("memprofile", "", "memory profile output file")
-	mountReadRetryTime = cmdMount2.Flag.Duration("readRetryTime", 6*time.Second, "maximum read retry wait time")
-}
-
-var cmdMount2 = &Command{
-	UsageLine: "mount2 -filer=localhost:8888 -dir=/some/dir",
-	Short:     "<WIP> mount weed filer to a directory as file system in userspace(FUSE)",
-	Long: `mount weed filer to userspace.
-
-  Pre-requisites:
-  1) have SeaweedFS master and volume servers running
-  2) have a "weed filer" running
-  These 2 requirements can be achieved with one command "weed server -filer=true"
-
-  This uses github.com/seaweedfs/fuse, which enables writing FUSE file systems on
-  Linux, and OS X.
-
-  On OS X, it requires OSXFUSE (http://osxfuse.github.com/).
-
-  `,
-}

weed/command/mount2_std.go

@@ -1,210 +0,0 @@
-package command
-
-import (
-	"context"
-	"fmt"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/mount"
-	"github.com/chrislusf/seaweedfs/weed/mount/meta_cache"
-	"github.com/chrislusf/seaweedfs/weed/mount/unmount"
-	"github.com/chrislusf/seaweedfs/weed/pb"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"github.com/chrislusf/seaweedfs/weed/security"
-	"github.com/chrislusf/seaweedfs/weed/storage/types"
-	"github.com/hanwen/go-fuse/v2/fuse"
-	"net/http"
-	"os"
-	"os/user"
-	"runtime"
-	"strconv"
-	"strings"
-	"time"
-
-	"github.com/chrislusf/seaweedfs/weed/util"
-	"github.com/chrislusf/seaweedfs/weed/util/grace"
-)
-
-func runMount2(cmd *Command, args []string) bool {
-
-	if *mountOptions.debug {
-		go http.ListenAndServe(fmt.Sprintf(":%d", *mountOptions.debugPort), nil)
-	}
-
-	grace.SetupProfiling(*mountCpuProfile, *mountMemProfile)
-	if *mountReadRetryTime < time.Second {
-		*mountReadRetryTime = time.Second
-	}
-	util.RetryWaitTime = *mountReadRetryTime
-
-	umask, umaskErr := strconv.ParseUint(*mountOptions.umaskString, 8, 64)
-	if umaskErr != nil {
-		fmt.Printf("can not parse umask %s", *mountOptions.umaskString)
-		return false
-	}
-
-	if len(args) > 0 {
-		return false
-	}
-
-	return RunMount2(&mount2Options, os.FileMode(umask))
-}
-
-func RunMount2(option *Mount2Options, umask os.FileMode) bool {
-
-	// basic checks
-	chunkSizeLimitMB := *mountOptions.chunkSizeLimitMB
-	if chunkSizeLimitMB <= 0 {
-		fmt.Printf("Please specify a reasonable buffer size.")
-		return false
-	}
-
-	// try to connect to filer
-	filerAddresses := pb.ServerAddresses(*option.filer).ToAddresses()
-	util.LoadConfiguration("security", false)
-	grpcDialOption := security.LoadClientTLS(util.GetViper(), "grpc.client")
-	var cipher bool
-	var err error
-	for i := 0; i < 10; i++ {
-		err = pb.WithOneOfGrpcFilerClients(false, filerAddresses, grpcDialOption, func(client filer_pb.SeaweedFilerClient) error {
-			resp, err := client.GetFilerConfiguration(context.Background(), &filer_pb.GetFilerConfigurationRequest{})
-			if err != nil {
-				return fmt.Errorf("get filer grpc address %v configuration: %v", filerAddresses, err)
-			}
-			cipher = resp.Cipher
-			return nil
-		})
-		if err != nil {
-			glog.V(0).Infof("failed to talk to filer %v: %v", filerAddresses, err)
-			glog.V(0).Infof("wait for %d seconds ...", i+1)
-			time.Sleep(time.Duration(i+1) * time.Second)
-		}
-	}
-	if err != nil {
-		glog.Errorf("failed to talk to filer %v: %v", filerAddresses, err)
-		return true
-	}
-
-	filerMountRootPath := *option.filerMountRootPath
-
-	// clean up mount point
-	dir := util.ResolvePath(*option.dir)
-	if dir == "" {
-		fmt.Printf("Please specify the mount directory via \"-dir\"")
-		return false
-	}
-
-	unmount.Unmount(dir)
-
-	// detect mount folder mode
-	if *option.dirAutoCreate {
-		os.MkdirAll(dir, os.FileMode(0777)&^umask)
-	}
-	fileInfo, err := os.Stat(dir)
-
-	// collect uid, gid
-	uid, gid := uint32(0), uint32(0)
-	mountMode := os.ModeDir | 0777
-	if err == nil {
-		mountMode = os.ModeDir | os.FileMode(0777)&^umask
-		uid, gid = util.GetFileUidGid(fileInfo)
-		fmt.Printf("mount point owner uid=%d gid=%d mode=%s\n", uid, gid, mountMode)
-	} else {
-		fmt.Printf("can not stat %s\n", dir)
-		return false
-	}
-
-	// detect uid, gid
-	if uid == 0 {
-		if u, err := user.Current(); err == nil {
-			if parsedId, pe := strconv.ParseUint(u.Uid, 10, 32); pe == nil {
-				uid = uint32(parsedId)
-			}
-			if parsedId, pe := strconv.ParseUint(u.Gid, 10, 32); pe == nil {
-				gid = uint32(parsedId)
-			}
-			fmt.Printf("current uid=%d gid=%d\n", uid, gid)
-		}
-	}
-
-	// mapping uid, gid
-	uidGidMapper, err := meta_cache.NewUidGidMapper(*option.uidMap, *option.gidMap)
-	if err != nil {
-		fmt.Printf("failed to parse %s %s: %v\n", *option.uidMap, *option.gidMap, err)
-		return false
-	}
-
-	// Ensure target mount point availability
-	if isValid := checkMountPointAvailable(dir); !isValid {
-		glog.Fatalf("Expected mount to still be active, target mount point: %s, please check!", dir)
-		return true
-	}
-
-	// mount fuse
-	fuseMountOptions := &fuse.MountOptions{
-		AllowOther:               *option.allowOthers,
-		Options:                  nil,
-		MaxBackground:            128,
-		MaxWrite:                 1024 * 1024 * 2,
-		MaxReadAhead:             1024 * 1024 * 2,
-		IgnoreSecurityLabels:     false,
-		RememberInodes:           false,
-		FsName:                   *option.filer + ":" + filerMountRootPath,
-		Name:                     "seaweedfs",
-		SingleThreaded:           false,
-		DisableXAttrs:            false,
-		Debug:                    true,
-		EnableLocks:              false,
-		ExplicitDataCacheControl: false,
-		// SyncRead:                 false, // set to false to enable the FUSE_CAP_ASYNC_READ capability
-		DirectMount:      true,
-		DirectMountFlags: 0,
-		// EnableAcl:                false,
-	}
-
-	// find mount point
-	mountRoot := filerMountRootPath
-	if mountRoot != "/" && strings.HasSuffix(mountRoot, "/") {
-		mountRoot = mountRoot[0 : len(mountRoot)-1]
-	}
-
-	seaweedFileSystem := mount.NewSeaweedFileSystem(&mount.Option{
-		MountDirectory:     dir,
-		FilerAddresses:     filerAddresses,
-		GrpcDialOption:     grpcDialOption,
-		FilerMountRootPath: mountRoot,
-		Collection:         *option.collection,
-		Replication:        *option.replication,
-		TtlSec:             int32(*option.ttlSec),
-		DiskType:           types.ToDiskType(*option.diskType),
-		ChunkSizeLimit:     int64(chunkSizeLimitMB) * 1024 * 1024,
-		ConcurrentWriters:  *option.concurrentWriters,
-		CacheDir:           *option.cacheDir,
-		CacheSizeMB:        *option.cacheSizeMB,
-		DataCenter:         *option.dataCenter,
-		MountUid:           uid,
-		MountGid:           gid,
-		MountMode:          mountMode,
-		MountCtime:         fileInfo.ModTime(),
-		MountMtime:         time.Now(),
-		Umask:              umask,
-		VolumeServerAccess: *mountOptions.volumeServerAccess,
-		Cipher:             cipher,
-		UidGidMapper:       uidGidMapper,
-	})
-
-	server, err := fuse.NewServer(seaweedFileSystem, dir, fuseMountOptions)
-	if err != nil {
-		glog.Fatalf("Mount fail: %v", err)
-	}
-	grace.OnInterrupt(func() {
-		unmount.Unmount(dir)
-	})
-
-	seaweedFileSystem.StartBackgroundTasks()
-
-	fmt.Printf("This is SeaweedFS version %s %s %s\n", util.Version(), runtime.GOOS, runtime.GOARCH)
-
-	server.Serve()
-
-	return true
-}

weed/command/mount_darwin.go

@@ -1,13 +1,5 @@
 package command
 
-import (
-	"github.com/seaweedfs/fuse"
-)
-
-func osSpecificMountOptions() []fuse.MountOption {
-	return []fuse.MountOption{}
-}
-
 func checkMountPointAvailable(dir string) bool {
 	return true
 }

weed/command/mount_freebsd.go

@@ -1,13 +0,0 @@
-package command
-
-import (
-	"github.com/seaweedfs/fuse"
-)
-
-func osSpecificMountOptions() []fuse.MountOption {
-	return []fuse.MountOption{}
-}
-
-func checkMountPointAvailable(dir string) bool {
-	return true
-}

weed/command/mount_linux.go

@@ -6,8 +6,6 @@ import (
 	"io"
 	"os"
 	"strings"
-
-	"github.com/seaweedfs/fuse"
 )
 
 const (
@@ -137,10 +135,6 @@ func parseInfoFile(r io.Reader) ([]*Info, error) {
 	return out, nil
 }
 
-func osSpecificMountOptions() []fuse.MountOption {
-	return []fuse.MountOption{}
-}
-
 func checkMountPointAvailable(dir string) bool {
 	mountPoint := dir
 	if mountPoint != "/" && strings.HasSuffix(mountPoint, "/") {

weed/command/mount_notsupported.go

@@ -1,5 +1,5 @@
-//go:build !linux && !darwin && !freebsd
-// +build !linux,!darwin,!freebsd
+//go:build !linux && !darwin
+// +build !linux,!darwin
 
 package command
 

weed/command/mount_std.go

@@ -1,34 +1,31 @@
-//go:build linux || darwin || freebsd
-// +build linux darwin freebsd
+//go:build linux || darwin
+// +build linux darwin
 
 package command
 
 import (
 	"context"
 	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/mount"
+	"github.com/chrislusf/seaweedfs/weed/mount/meta_cache"
+	"github.com/chrislusf/seaweedfs/weed/mount/unmount"
+	"github.com/chrislusf/seaweedfs/weed/pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/pb/mount_pb"
+	"github.com/chrislusf/seaweedfs/weed/security"
+	"github.com/chrislusf/seaweedfs/weed/storage/types"
+	"github.com/hanwen/go-fuse/v2/fuse"
+	"google.golang.org/grpc/reflection"
+	"net"
 	"net/http"
 	"os"
 	"os/user"
-	"path"
-	"path/filepath"
 	"runtime"
 	"strconv"
 	"strings"
-	"syscall"
 	"time"
 
-	"github.com/chrislusf/seaweedfs/weed/storage/types"
-
-	"github.com/chrislusf/seaweedfs/weed/filesys/meta_cache"
-
-	"github.com/seaweedfs/fuse"
-	"github.com/seaweedfs/fuse/fs"
-
-	"github.com/chrislusf/seaweedfs/weed/filesys"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/pb"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"github.com/chrislusf/seaweedfs/weed/security"
 	"github.com/chrislusf/seaweedfs/weed/util"
 	"github.com/chrislusf/seaweedfs/weed/util/grace"
 )
@@ -58,27 +55,18 @@ func runMount(cmd *Command, args []string) bool {
 	return RunMount(&mountOptions, os.FileMode(umask))
 }
 
-func getParentInode(mountDir string) (uint64, error) {
-	parentDir := filepath.Clean(filepath.Join(mountDir, ".."))
-	fi, err := os.Stat(parentDir)
-	if err != nil {
-		return 0, err
-	}
-
-	stat, ok := fi.Sys().(*syscall.Stat_t)
-	if !ok {
-		return 0, nil
-	}
-
-	return stat.Ino, nil
-}
-
 func RunMount(option *MountOptions, umask os.FileMode) bool {
 
-	filerAddresses := pb.ServerAddresses(*option.filer).ToAddresses()
+	// basic checks
+	chunkSizeLimitMB := *mountOptions.chunkSizeLimitMB
+	if chunkSizeLimitMB <= 0 {
+		fmt.Printf("Please specify a reasonable buffer size.")
+		return false
+	}
 
+	// try to connect to filer
+	filerAddresses := pb.ServerAddresses(*option.filer).ToAddresses()
 	util.LoadConfiguration("security", false)
-	// try to connect to filer, filerBucketsPath may be useful later
 	grpcDialOption := security.LoadClientTLS(util.GetViper(), "grpc.client")
 	var cipher bool
 	var err error
@@ -103,26 +91,31 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 	}
 
 	filerMountRootPath := *option.filerMountRootPath
-	dir := util.ResolvePath(*option.dir)
-	parentInode, err := getParentInode(dir)
-	if err != nil {
-		glog.Errorf("failed to retrieve inode for parent directory of %s: %v", dir, err)
-		return true
-	}
 
-	fmt.Printf("This is SeaweedFS version %s %s %s\n", util.Version(), runtime.GOOS, runtime.GOARCH)
+	// clean up mount point
+	dir := util.ResolvePath(*option.dir)
 	if dir == "" {
 		fmt.Printf("Please specify the mount directory via \"-dir\"")
 		return false
 	}
 
-	chunkSizeLimitMB := *mountOptions.chunkSizeLimitMB
-	if chunkSizeLimitMB <= 0 {
-		fmt.Printf("Please specify a reasonable buffer size.")
-		return false
-	}
+	unmount.Unmount(dir)
 
-	fuse.Unmount(dir)
+	// start on local unix socket
+	if *option.localSocket == "" {
+		mountDirHash := util.HashToInt32([]byte(dir))
+		if mountDirHash < 0 {
+			mountDirHash = -mountDirHash
+		}
+		*option.localSocket = fmt.Sprintf("/tmp/seaweefs-mount-%d.sock", mountDirHash)
+	}
+	if err := os.Remove(*option.localSocket); err != nil && !os.IsNotExist(err) {
+		glog.Fatalf("Failed to remove %s, error: %s", *option.localSocket, err.Error())
+	}
+	montSocketListener, err := net.Listen("unix", *option.localSocket)
+	if err != nil {
+		glog.Fatalf("Failed to listen on %s: %v", *option.localSocket, err)
+	}
 
 	// detect mount folder mode
 	if *option.dirAutoCreate {
@@ -130,6 +123,7 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 	}
 	fileInfo, err := os.Stat(dir)
 
+	// collect uid, gid
 	uid, gid := uint32(0), uint32(0)
 	mountMode := os.ModeDir | 0777
 	if err == nil {
@@ -141,6 +135,7 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 		return false
 	}
 
+	// detect uid, gid
 	if uid == 0 {
 		if u, err := user.Current(); err == nil {
 			if parsedId, pe := strconv.ParseUint(u.Uid, 10, 32); pe == nil {
@@ -166,35 +161,51 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 		return true
 	}
 
-	mountName := path.Base(dir)
+	serverFriendlyName := strings.ReplaceAll(*option.filer, ",", "+")
 
-	options := []fuse.MountOption{
-		fuse.VolumeName(mountName),
-		fuse.FSName(*option.filer + ":" + filerMountRootPath),
-		fuse.Subtype("seaweedfs"),
-		// fuse.NoAppleDouble(), // include .DS_Store, otherwise can not delete non-empty folders
-		fuse.NoAppleXattr(),
-		fuse.ExclCreate(),
-		fuse.DaemonTimeout("3600"),
-		fuse.AllowDev(),
-		fuse.AllowSUID(),
-		fuse.DefaultPermissions(),
-		fuse.MaxReadahead(1024 * 512),
-		fuse.AsyncRead(),
-		// fuse.WritebackCache(),
-		// fuse.MaxBackground(1024),
-		// fuse.CongestionThreshold(1024),
-	}
-
-	options = append(options, osSpecificMountOptions()...)
-	if *option.allowOthers {
-		options = append(options, fuse.AllowOther())
+	// mount fuse
+	fuseMountOptions := &fuse.MountOptions{
+		AllowOther:               *option.allowOthers,
+		Options:                  nil,
+		MaxBackground:            128,
+		MaxWrite:                 1024 * 1024 * 2,
+		MaxReadAhead:             1024 * 1024 * 2,
+		IgnoreSecurityLabels:     false,
+		RememberInodes:           false,
+		FsName:                   serverFriendlyName + ":" + filerMountRootPath,
+		Name:                     "seaweedfs",
+		SingleThreaded:           false,
+		DisableXAttrs:            false,
+		Debug:                    *option.debug,
+		EnableLocks:              false,
+		ExplicitDataCacheControl: false,
+		DirectMount:              true,
+		DirectMountFlags:         0,
+		//SyncRead:                 false, // set to false to enable the FUSE_CAP_ASYNC_READ capability
+		//EnableAcl:                true,
 	}
 	if *option.nonempty {
-		options = append(options, fuse.AllowNonEmptyMount())
+		fuseMountOptions.Options = append(fuseMountOptions.Options, "nonempty")
 	}
 	if *option.readOnly {
-		options = append(options, fuse.ReadOnly())
+		if runtime.GOOS == "darwin" {
+			fuseMountOptions.Options = append(fuseMountOptions.Options, "rdonly")
+		} else {
+			fuseMountOptions.Options = append(fuseMountOptions.Options, "ro")
+		}
+	}
+	if runtime.GOOS == "darwin" {
+		// https://github-wiki-see.page/m/macfuse/macfuse/wiki/Mount-Options
+		ioSizeMB := 1
+		for ioSizeMB*2 <= *option.chunkSizeLimitMB && ioSizeMB*2 <= 32 {
+			ioSizeMB *= 2
+		}
+		fuseMountOptions.Options = append(fuseMountOptions.Options, "daemon_timeout=600")
+		fuseMountOptions.Options = append(fuseMountOptions.Options, "noapplexattr")
+		// fuseMountOptions.Options = append(fuseMountOptions.Options, "novncache") // need to test effectiveness
+		fuseMountOptions.Options = append(fuseMountOptions.Options, "slow_statfs")
+		fuseMountOptions.Options = append(fuseMountOptions.Options, "volname="+serverFriendlyName)
+		fuseMountOptions.Options = append(fuseMountOptions.Options, fmt.Sprintf("iosize=%d", ioSizeMB*1024*1024))
 	}
 
 	// find mount point
@@ -203,9 +214,7 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 		mountRoot = mountRoot[0 : len(mountRoot)-1]
 	}
 
-	diskType := types.ToDiskType(*option.diskType)
-
-	seaweedFileSystem := filesys.NewSeaweedFileSystem(&filesys.Option{
+	seaweedFileSystem := mount.NewSeaweedFileSystem(&mount.Option{
 		MountDirectory:     dir,
 		FilerAddresses:     filerAddresses,
 		GrpcDialOption:     grpcDialOption,
@@ -213,49 +222,42 @@ func RunMount(option *MountOptions, umask os.FileMode) bool {
 		Collection:         *option.collection,
 		Replication:        *option.replication,
 		TtlSec:             int32(*option.ttlSec),
-		DiskType:           diskType,
+		DiskType:           types.ToDiskType(*option.diskType),
 		ChunkSizeLimit:     int64(chunkSizeLimitMB) * 1024 * 1024,
 		ConcurrentWriters:  *option.concurrentWriters,
 		CacheDir:           *option.cacheDir,
 		CacheSizeMB:        *option.cacheSizeMB,
 		DataCenter:         *option.dataCenter,
+		Quota:              int64(*option.collectionQuota) * 1024 * 1024,
 		MountUid:           uid,
 		MountGid:           gid,
 		MountMode:          mountMode,
 		MountCtime:         fileInfo.ModTime(),
 		MountMtime:         time.Now(),
-		MountParentInode:   parentInode,
 		Umask:              umask,
 		VolumeServerAccess: *mountOptions.volumeServerAccess,
 		Cipher:             cipher,
 		UidGidMapper:       uidGidMapper,
 	})
 
-	// mount
-	c, err := fuse.Mount(dir, options...)
+	server, err := fuse.NewServer(seaweedFileSystem, dir, fuseMountOptions)
 	if err != nil {
-		glog.V(0).Infof("mount: %v", err)
-		return true
+		glog.Fatalf("Mount fail: %v", err)
 	}
-	defer fuse.Unmount(dir)
-
 	grace.OnInterrupt(func() {
-		fuse.Unmount(dir)
-		c.Close()
+		unmount.Unmount(dir)
 	})
 
-	glog.V(0).Infof("mounted %s%s to %v", *option.filer, mountRoot, dir)
-	server := fs.New(c, nil)
-	seaweedFileSystem.Server = server
-	seaweedFileSystem.StartBackgroundTasks()
-	err = server.Serve(seaweedFileSystem)
+	grpcS := pb.NewGrpcServer()
+	mount_pb.RegisterSeaweedMountServer(grpcS, seaweedFileSystem)
+	reflection.Register(grpcS)
+	go grpcS.Serve(montSocketListener)
 
-	// check if the mount process has an error to report
-	<-c.Ready
-	if err := c.MountError; err != nil {
-		glog.V(0).Infof("mount process: %v", err)
-		return true
-	}
+	seaweedFileSystem.StartBackgroundTasks()
+
+	fmt.Printf("This is SeaweedFS version %s %s %s\n", util.Version(), runtime.GOOS, runtime.GOARCH)
+
+	server.Serve()
 
 	return true
 }

weed/command/msg_broker.go

@@ -95,7 +95,7 @@ func (msgBrokerOpt *MessageBrokerOptions) startQueueServer() bool {
 	}, grpcDialOption)
 
 	// start grpc listener
-	grpcL, err := util.NewListener(util.JoinHostPort("", *msgBrokerOpt.port), 0)
+	grpcL, _, err := util.NewIpAndLocalListeners("", *msgBrokerOpt.port, 0)
 	if err != nil {
 		glog.Fatalf("failed to listen on grpc port %d: %v", *msgBrokerOpt.port, err)
 	}

weed/command/s3.go

@@ -24,22 +24,24 @@ var (
 )
 
 type S3Options struct {
-	filer            *string
-	bindIp           *string
-	port             *int
-	config           *string
-	domainName       *string
-	tlsPrivateKey    *string
-	tlsCertificate   *string
-	metricsHttpPort  *int
-	allowEmptyFolder *bool
-	auditLogConfig   *string
+	filer                     *string
+	bindIp                    *string
+	port                      *int
+	config                    *string
+	domainName                *string
+	tlsPrivateKey             *string
+	tlsCertificate            *string
+	metricsHttpPort           *int
+	allowEmptyFolder          *bool
+	allowDeleteBucketNotEmpty *bool
+	auditLogConfig            *string
+	localFilerSocket          *string
 }
 
 func init() {
 	cmdS3.Run = runS3 // break init cycle
 	s3StandaloneOptions.filer = cmdS3.Flag.String("filer", "localhost:8888", "filer server address")
-	s3StandaloneOptions.bindIp = cmdS3.Flag.String("ip.bind", "", "ip address to bind to")
+	s3StandaloneOptions.bindIp = cmdS3.Flag.String("ip.bind", "", "ip address to bind to. Default to localhost.")
 	s3StandaloneOptions.port = cmdS3.Flag.Int("port", 8333, "s3 server http listen port")
 	s3StandaloneOptions.domainName = cmdS3.Flag.String("domainName", "", "suffix of the host name in comma separated list, {bucket}.{domainName}")
 	s3StandaloneOptions.config = cmdS3.Flag.String("config", "", "path to the config file")
@@ -48,6 +50,7 @@ func init() {
 	s3StandaloneOptions.tlsCertificate = cmdS3.Flag.String("cert.file", "", "path to the TLS certificate file")
 	s3StandaloneOptions.metricsHttpPort = cmdS3.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
 	s3StandaloneOptions.allowEmptyFolder = cmdS3.Flag.Bool("allowEmptyFolder", true, "allow empty folders")
+	s3StandaloneOptions.allowDeleteBucketNotEmpty = cmdS3.Flag.Bool("allowDeleteBucketNotEmpty", true, "allow recursive deleting all entries along with bucket")
 }
 
 var cmdS3 = &Command{
@@ -177,13 +180,15 @@ func (s3opt *S3Options) startS3Server() bool {
 	router := mux.NewRouter().SkipClean(true)
 
 	_, s3ApiServer_err := s3api.NewS3ApiServer(router, &s3api.S3ApiServerOption{
-		Filer:            filerAddress,
-		Port:             *s3opt.port,
-		Config:           *s3opt.config,
-		DomainName:       *s3opt.domainName,
-		BucketsPath:      filerBucketsPath,
-		GrpcDialOption:   grpcDialOption,
-		AllowEmptyFolder: *s3opt.allowEmptyFolder,
+		Filer:                     filerAddress,
+		Port:                      *s3opt.port,
+		Config:                    *s3opt.config,
+		DomainName:                *s3opt.domainName,
+		BucketsPath:               filerBucketsPath,
+		GrpcDialOption:            grpcDialOption,
+		AllowEmptyFolder:          *s3opt.allowEmptyFolder,
+		AllowDeleteBucketNotEmpty: *s3opt.allowDeleteBucketNotEmpty,
+		LocalFilerSocket:          s3opt.localFilerSocket,
 	})
 	if s3ApiServer_err != nil {
 		glog.Fatalf("S3 API Server startup error: %v", s3ApiServer_err)
@@ -191,8 +196,12 @@ func (s3opt *S3Options) startS3Server() bool {
 
 	httpS := &http.Server{Handler: router}
 
+	if *s3opt.bindIp == "" {
+		*s3opt.bindIp = "localhost"
+	}
+
 	listenAddress := fmt.Sprintf("%s:%d", *s3opt.bindIp, *s3opt.port)
-	s3ApiListener, err := util.NewListener(listenAddress, time.Duration(10)*time.Second)
+	s3ApiListener, s3ApiLocalListner, err := util.NewIpAndLocalListeners(*s3opt.bindIp, *s3opt.port, time.Duration(10)*time.Second)
 	if err != nil {
 		glog.Fatalf("S3 API Server listener on %s error: %v", listenAddress, err)
 	}
@@ -206,11 +215,25 @@ func (s3opt *S3Options) startS3Server() bool {
 
 	if *s3opt.tlsPrivateKey != "" {
 		glog.V(0).Infof("Start Seaweed S3 API Server %s at https port %d", util.Version(), *s3opt.port)
+		if s3ApiLocalListner != nil {
+			go func() {
+				if err = httpS.ServeTLS(s3ApiLocalListner, *s3opt.tlsCertificate, *s3opt.tlsPrivateKey); err != nil {
+					glog.Fatalf("S3 API Server Fail to serve: %v", err)
+				}
+			}()
+		}
 		if err = httpS.ServeTLS(s3ApiListener, *s3opt.tlsCertificate, *s3opt.tlsPrivateKey); err != nil {
 			glog.Fatalf("S3 API Server Fail to serve: %v", err)
 		}
 	} else {
 		glog.V(0).Infof("Start Seaweed S3 API Server %s at http port %d", util.Version(), *s3opt.port)
+		if s3ApiLocalListner != nil {
+			go func() {
+				if err = httpS.Serve(s3ApiLocalListner); err != nil {
+					glog.Fatalf("S3 API Server Fail to serve: %v", err)
+				}
+			}()
+		}
 		if err = httpS.Serve(s3ApiListener); err != nil {
 			glog.Fatalf("S3 API Server Fail to serve: %v", err)
 		}

weed/command/scaffold/filer.toml

@@ -195,6 +195,40 @@ routeByLatency = false
 # This changes the data layout. Only add new directories. Removing/Updating will cause data loss.
 superLargeDirectories = []
 
+[redis_lua]
+enabled = false
+address = "localhost:6379"
+password = ""
+database = 0
+# This changes the data layout. Only add new directories. Removing/Updating will cause data loss.
+superLargeDirectories = []
+
+[redis_lua_sentinel]
+enabled = false
+addresses = ["172.22.12.7:26379","172.22.12.8:26379","172.22.12.9:26379"]
+masterName = "master"
+username = ""
+password = ""
+database = 0
+
+[redis_lua_cluster]
+enabled = false
+addresses = [
+    "localhost:30001",
+    "localhost:30002",
+    "localhost:30003",
+    "localhost:30004",
+    "localhost:30005",
+    "localhost:30006",
+]
+password = ""
+# allows reads from slave servers or the master, but all writes still go to the master
+readOnly = false
+# automatically use the closest Redis server for reads
+routeByLatency = false
+# This changes the data layout. Only add new directories. Removing/Updating will cause data loss.
+superLargeDirectories = []
+
 [redis3] # beta
 enabled = false
 address = "localhost:6379"

weed/command/scaffold/security.toml

@@ -83,7 +83,13 @@ key = ""
 #     this does not work with other clients, e.g., "weed filer|mount" etc, yet.
 [https.client]
 enabled = true
+
 [https.volume]
 cert = ""
 key = ""
+ca = ""
 
+[https.master]
+cert = ""
+key = ""
+ca = ""

weed/command/server.go

@@ -55,7 +55,7 @@ var cmdServer = &Command{
 
 var (
 	serverIp                  = cmdServer.Flag.String("ip", util.DetectedHostAddress(), "ip or server name, also used as identifier")
-	serverBindIp              = cmdServer.Flag.String("ip.bind", "", "ip address to bind to")
+	serverBindIp              = cmdServer.Flag.String("ip.bind", "", "ip address to bind to. If empty, default to same as -ip option.")
 	serverTimeout             = cmdServer.Flag.Int("idleTimeout", 30, "connection idle seconds")
 	serverDataCenter          = cmdServer.Flag.String("dataCenter", "", "current volume server's data center name")
 	serverRack                = cmdServer.Flag.String("rack", "", "current volume server's rack name")
@@ -98,6 +98,8 @@ func init() {
 	masterOptions.metricsAddress = cmdServer.Flag.String("metrics.address", "", "Prometheus gateway address")
 	masterOptions.metricsIntervalSec = cmdServer.Flag.Int("metrics.intervalSeconds", 15, "Prometheus push interval in seconds")
 	masterOptions.raftResumeState = cmdServer.Flag.Bool("resumeState", false, "resume previous state on start master server")
+	masterOptions.heartbeatInterval = cmdServer.Flag.Duration("master.heartbeatInterval", 300*time.Millisecond, "heartbeat interval of master servers, and will be randomly multiplied by [1, 1.25)")
+	masterOptions.electionTimeout = cmdServer.Flag.Duration("master.electionTimeout", 10*time.Second, "election timeout of master servers")
 
 	filerOptions.collection = cmdServer.Flag.String("filer.collection", "", "all data will be stored in this collection")
 	filerOptions.port = cmdServer.Flag.Int("filer.port", 8888, "filer server http listen port")
@@ -110,6 +112,7 @@ func init() {
 	filerOptions.cipher = cmdServer.Flag.Bool("filer.encryptVolumeData", false, "encrypt data on volume servers")
 	filerOptions.saveToFilerLimit = cmdServer.Flag.Int("filer.saveToFilerLimit", 0, "Small files smaller than this limit can be cached in filer store.")
 	filerOptions.concurrentUploadLimitMB = cmdServer.Flag.Int("filer.concurrentUploadLimitMB", 64, "limit total concurrent upload size")
+	filerOptions.localSocket = cmdServer.Flag.String("filer.localSocket", "", "default to /tmp/seaweedfs-filer-<port>.sock")
 
 	serverOptions.v.port = cmdServer.Flag.Int("volume.port", 8080, "volume server http listen port")
 	serverOptions.v.portGrpc = cmdServer.Flag.Int("volume.port.grpc", 0, "volume server grpc listen port")
@@ -135,6 +138,7 @@ func init() {
 	s3Options.config = cmdServer.Flag.String("s3.config", "", "path to the config file")
 	s3Options.auditLogConfig = cmdServer.Flag.String("s3.auditLogConfig", "", "path to the audit log config file")
 	s3Options.allowEmptyFolder = cmdServer.Flag.Bool("s3.allowEmptyFolder", true, "allow empty folders")
+	s3Options.allowDeleteBucketNotEmpty = cmdServer.Flag.Bool("s3.allowDeleteBucketNotEmpty", true, "allow recursive deleting all entries along with bucket")
 
 	iamOptions.port = cmdServer.Flag.Int("iam.port", 8111, "iam server http listen port")
 
@@ -145,7 +149,7 @@ func init() {
 	webdavOptions.tlsPrivateKey = cmdServer.Flag.String("webdav.key.file", "", "path to the TLS private key file")
 	webdavOptions.tlsCertificate = cmdServer.Flag.String("webdav.cert.file", "", "path to the TLS certificate file")
 	webdavOptions.cacheDir = cmdServer.Flag.String("webdav.cacheDir", os.TempDir(), "local cache directory for file chunks")
-	webdavOptions.cacheSizeMB = cmdServer.Flag.Int64("webdav.cacheCapacityMB", 1000, "local cache capacity in MB")
+	webdavOptions.cacheSizeMB = cmdServer.Flag.Int64("webdav.cacheCapacityMB", 0, "local cache capacity in MB")
 
 	msgBrokerOptions.port = cmdServer.Flag.Int("msgBroker.port", 17777, "broker gRPC listen port")
 
@@ -181,13 +185,19 @@ func runServer(cmd *Command, args []string) bool {
 		masterOptions.peers = &peers
 	}
 
+	if *serverBindIp == "" {
+		serverBindIp = serverIp
+	}
+
 	// ip address
 	masterOptions.ip = serverIp
 	masterOptions.ipBind = serverBindIp
-	filerOptions.masters = pb.ServerAddresses(*masterOptions.peers).ToAddresses()
+	filerOptions.masters = pb.ServerAddresses(*masterOptions.peers).ToAddressMap()
 	filerOptions.ip = serverIp
 	filerOptions.bindIp = serverBindIp
 	s3Options.bindIp = serverBindIp
+	iamOptions.ip = serverBindIp
+	iamOptions.masters = masterOptions.peers
 	serverOptions.v.ip = serverIp
 	serverOptions.v.bindIp = serverBindIp
 	serverOptions.v.masters = pb.ServerAddresses(*masterOptions.peers).ToAddresses()
@@ -242,6 +252,7 @@ func runServer(cmd *Command, args []string) bool {
 	if *isStartingS3 {
 		go func() {
 			time.Sleep(2 * time.Second)
+			s3Options.localFilerSocket = filerOptions.localSocket
 			s3Options.startS3Server()
 		}()
 	}

weed/command/volume.go

@@ -2,7 +2,6 @@ package command
 
 import (
 	"fmt"
-	"github.com/chrislusf/seaweedfs/weed/storage/types"
 	"net/http"
 	httppprof "net/http/pprof"
 	"os"
@@ -11,6 +10,8 @@ import (
 	"strings"
 	"time"
 
+	"github.com/chrislusf/seaweedfs/weed/storage/types"
+
 	"github.com/spf13/viper"
 	"google.golang.org/grpc"
 
@@ -24,7 +25,7 @@ import (
 
 	"github.com/chrislusf/seaweedfs/weed/glog"
 	"github.com/chrislusf/seaweedfs/weed/pb/volume_server_pb"
-	"github.com/chrislusf/seaweedfs/weed/server"
+	weed_server "github.com/chrislusf/seaweedfs/weed/server"
 	stats_collect "github.com/chrislusf/seaweedfs/weed/stats"
 	"github.com/chrislusf/seaweedfs/weed/storage"
 	"github.com/chrislusf/seaweedfs/weed/util"
@@ -74,7 +75,7 @@ func init() {
 	v.publicPort = cmdVolume.Flag.Int("port.public", 0, "port opened to public")
 	v.ip = cmdVolume.Flag.String("ip", util.DetectedHostAddress(), "ip or server name, also used as identifier")
 	v.publicUrl = cmdVolume.Flag.String("publicUrl", "", "Publicly accessible address")
-	v.bindIp = cmdVolume.Flag.String("ip.bind", "", "ip address to bind to")
+	v.bindIp = cmdVolume.Flag.String("ip.bind", "", "ip address to bind to. If empty, default to same as -ip option.")
 	v.mastersString = cmdVolume.Flag.String("mserver", "localhost:9333", "comma-separated master servers")
 	v.preStopSeconds = cmdVolume.Flag.Int("preStopSeconds", 10, "number of seconds between stop send heartbeats and stop volume server")
 	// v.pulseSeconds = cmdVolume.Flag.Int("pulseSeconds", 5, "number of seconds between heartbeats, must be smaller than or equal to the master's setting")
@@ -94,7 +95,7 @@ func init() {
 	v.pprof = cmdVolume.Flag.Bool("pprof", false, "enable pprof http handlers. precludes --memprofile and --cpuprofile")
 	v.metricsHttpPort = cmdVolume.Flag.Int("metricsPort", 0, "Prometheus metrics listen port")
 	v.idxFolder = cmdVolume.Flag.String("dir.idx", "", "directory to store .idx files")
-	v.enableTcp = cmdVolume.Flag.Bool("tcp", false, "<exprimental> enable tcp port")
+	v.enableTcp = cmdVolume.Flag.Bool("tcp", false, "<experimental> enable tcp port")
 }
 
 var cmdVolume = &Command{
@@ -193,6 +194,9 @@ func (v VolumeServerOptions) startVolumeServer(volumeFolders, maxVolumeCounts, v
 		*v.ip = util.DetectedHostAddress()
 		glog.V(0).Infof("detected volume server ip address: %v", *v.ip)
 	}
+	if *v.bindIp == "" {
+		*v.bindIp = *v.ip
+	}
 
 	if *v.publicPort == 0 {
 		*v.publicPort = *v.port
@@ -364,11 +368,18 @@ func (v VolumeServerOptions) startClusterHttpService(handler http.Handler) httpd
 	}
 
 	httpDown := httpdown.HTTP{
-		KillTimeout: 5 * time.Minute,
-		StopTimeout: 5 * time.Minute,
+		KillTimeout: time.Minute,
+		StopTimeout: 30 * time.Second,
 		CertFile:    certFile,
 		KeyFile:     keyFile}
-	clusterHttpServer := httpDown.Serve(&http.Server{Handler: handler}, listener)
+	httpS := &http.Server{Handler: handler}
+
+	if viper.GetString("https.volume.ca") != "" {
+		clientCertFile := viper.GetString("https.volume.ca")
+		httpS.TLSConfig = security.LoadClientTLSHTTP(clientCertFile)
+	}
+
+	clusterHttpServer := httpDown.Serve(httpS, listener)
 	go func() {
 		if e := clusterHttpServer.Wait(); e != nil {
 			glog.Fatalf("Volume server fail to serve: %v", e)

weed/command/webdav.go

@@ -43,7 +43,7 @@ func init() {
 	webDavStandaloneOptions.tlsPrivateKey = cmdWebDav.Flag.String("key.file", "", "path to the TLS private key file")
 	webDavStandaloneOptions.tlsCertificate = cmdWebDav.Flag.String("cert.file", "", "path to the TLS certificate file")
 	webDavStandaloneOptions.cacheDir = cmdWebDav.Flag.String("cacheDir", os.TempDir(), "local cache directory for file chunks")
-	webDavStandaloneOptions.cacheSizeMB = cmdWebDav.Flag.Int64("cacheCapacityMB", 1000, "local cache capacity in MB")
+	webDavStandaloneOptions.cacheSizeMB = cmdWebDav.Flag.Int64("cacheCapacityMB", 0, "local cache capacity in MB")
 }
 
 var cmdWebDav = &Command{

weed/filer/entry.go

@@ -24,6 +24,8 @@ type Attr struct {
 	SymlinkTarget string
 	Md5           []byte
 	FileSize      uint64
+	Rdev          uint32
+	Inode         uint64
 }
 
 func (attr Attr) IsDirectory() bool {

weed/filer/entry_codec.go

@@ -48,6 +48,8 @@ func EntryAttributeToPb(entry *Entry) *filer_pb.FuseAttributes {
 		SymlinkTarget: entry.Attr.SymlinkTarget,
 		Md5:           entry.Attr.Md5,
 		FileSize:      entry.Attr.FileSize,
+		Rdev:          entry.Attr.Rdev,
+		Inode:         entry.Attr.Inode,
 	}
 }
 
@@ -74,6 +76,8 @@ func PbToEntryAttribute(attr *filer_pb.FuseAttributes) Attr {
 	t.SymlinkTarget = attr.SymlinkTarget
 	t.Md5 = attr.Md5
 	t.FileSize = attr.FileSize
+	t.Rdev = attr.Rdev
+	t.Inode = attr.Inode
 
 	return t
 }

weed/filer/filechunk_manifest.go

@@ -8,6 +8,7 @@ import (
 	"math"
 	"net/url"
 	"strings"
+	"sync"
 	"time"
 
 	"github.com/golang/protobuf/proto"
@@ -21,6 +22,12 @@ const (
 	ManifestBatch = 10000
 )
 
+var bytesBufferPool = sync.Pool{
+	New: func() interface{} {
+		return new(bytes.Buffer)
+	},
+}
+
 func HasChunkManifest(chunks []*filer_pb.FileChunk) bool {
 	for _, chunk := range chunks {
 		if chunk.IsChunkManifest {
@@ -61,12 +68,12 @@ func ResolveChunkManifest(lookupFileIdFn wdclient.LookupFileIdFunctionType, chun
 
 		manifestChunks = append(manifestChunks, chunk)
 		// recursive
-		dchunks, mchunks, subErr := ResolveChunkManifest(lookupFileIdFn, resolvedChunks, startOffset, stopOffset)
+		dataChunks, manifestChunks, subErr := ResolveChunkManifest(lookupFileIdFn, resolvedChunks, startOffset, stopOffset)
 		if subErr != nil {
 			return chunks, nil, subErr
 		}
-		dataChunks = append(dataChunks, dchunks...)
-		manifestChunks = append(manifestChunks, mchunks...)
+		dataChunks = append(dataChunks, dataChunks...)
+		manifestChunks = append(manifestChunks, manifestChunks...)
 	}
 	return
 }
@@ -77,12 +84,15 @@ func ResolveOneChunkManifest(lookupFileIdFn wdclient.LookupFileIdFunctionType, c
 	}
 
 	// IsChunkManifest
-	data, err := fetchChunk(lookupFileIdFn, chunk.GetFileIdString(), chunk.CipherKey, chunk.IsCompressed)
+	bytesBuffer := bytesBufferPool.Get().(*bytes.Buffer)
+	bytesBuffer.Reset()
+	defer bytesBufferPool.Put(bytesBuffer)
+	err := fetchWholeChunk(bytesBuffer, lookupFileIdFn, chunk.GetFileIdString(), chunk.CipherKey, chunk.IsCompressed)
 	if err != nil {
 		return nil, fmt.Errorf("fail to read manifest %s: %v", chunk.GetFileIdString(), err)
 	}
 	m := &filer_pb.FileChunkManifest{}
-	if err := proto.Unmarshal(data, m); err != nil {
+	if err := proto.Unmarshal(bytesBuffer.Bytes(), m); err != nil {
 		return nil, fmt.Errorf("fail to unmarshal manifest %s: %v", chunk.GetFileIdString(), err)
 	}
 
@@ -92,38 +102,43 @@ func ResolveOneChunkManifest(lookupFileIdFn wdclient.LookupFileIdFunctionType, c
 }
 
 // TODO fetch from cache for weed mount?
-func fetchChunk(lookupFileIdFn wdclient.LookupFileIdFunctionType, fileId string, cipherKey []byte, isGzipped bool) ([]byte, error) {
+func fetchWholeChunk(bytesBuffer *bytes.Buffer, lookupFileIdFn wdclient.LookupFileIdFunctionType, fileId string, cipherKey []byte, isGzipped bool) error {
 	urlStrings, err := lookupFileIdFn(fileId)
 	if err != nil {
 		glog.Errorf("operation LookupFileId %s failed, err: %v", fileId, err)
-		return nil, err
+		return err
 	}
-	return retriedFetchChunkData(urlStrings, cipherKey, isGzipped, true, 0, 0)
+	err = retriedStreamFetchChunkData(bytesBuffer, urlStrings, cipherKey, isGzipped, true, 0, 0)
+	if err != nil {
+		return err
+	}
+	return nil
 }
 
-func fetchChunkRange(lookupFileIdFn wdclient.LookupFileIdFunctionType, fileId string, cipherKey []byte, isGzipped bool, offset int64, size int) ([]byte, error) {
+func fetchChunkRange(buffer []byte, lookupFileIdFn wdclient.LookupFileIdFunctionType, fileId string, cipherKey []byte, isGzipped bool, offset int64) (int, error) {
 	urlStrings, err := lookupFileIdFn(fileId)
 	if err != nil {
 		glog.Errorf("operation LookupFileId %s failed, err: %v", fileId, err)
-		return nil, err
+		return 0, err
 	}
-	return retriedFetchChunkData(urlStrings, cipherKey, isGzipped, false, offset, size)
+	return retriedFetchChunkData(buffer, urlStrings, cipherKey, isGzipped, false, offset)
 }
 
-func retriedFetchChunkData(urlStrings []string, cipherKey []byte, isGzipped bool, isFullChunk bool, offset int64, size int) ([]byte, error) {
+func retriedFetchChunkData(buffer []byte, urlStrings []string, cipherKey []byte, isGzipped bool, isFullChunk bool, offset int64) (n int, err error) {
 
-	var err error
 	var shouldRetry bool
-	receivedData := make([]byte, 0, size)
 
 	for waitTime := time.Second; waitTime < util.RetryWaitTime; waitTime += waitTime / 2 {
 		for _, urlString := range urlStrings {
-			receivedData = receivedData[:0]
+			n = 0
 			if strings.Contains(urlString, "%") {
 				urlString = url.PathEscape(urlString)
 			}
-			shouldRetry, err = util.ReadUrlAsStream(urlString+"?readDeleted=true", cipherKey, isGzipped, isFullChunk, offset, size, func(data []byte) {
-				receivedData = append(receivedData, data...)
+			shouldRetry, err = util.ReadUrlAsStream(urlString+"?readDeleted=true", cipherKey, isGzipped, isFullChunk, offset, len(buffer), func(data []byte) {
+				if n < len(buffer) {
+					x := copy(buffer[n:], data)
+					n += x
+				}
 			})
 			if !shouldRetry {
 				break
@@ -142,7 +157,7 @@ func retriedFetchChunkData(urlStrings []string, cipherKey []byte, isGzipped bool
 		}
 	}
 
-	return receivedData, err
+	return n, err
 
 }
 

weed/filer/filechunks.go

@@ -4,8 +4,8 @@ import (
 	"bytes"
 	"fmt"
 	"github.com/chrislusf/seaweedfs/weed/wdclient"
+	"golang.org/x/exp/slices"
 	"math"
-	"sort"
 	"sync"
 
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
@@ -23,7 +23,16 @@ func TotalSize(chunks []*filer_pb.FileChunk) (size uint64) {
 }
 
 func FileSize(entry *filer_pb.Entry) (size uint64) {
-	return maxUint64(TotalSize(entry.Chunks), entry.Attributes.FileSize)
+	if entry == nil || entry.Attributes == nil {
+		return 0
+	}
+	fileSize := entry.Attributes.FileSize
+	if entry.RemoteEntry != nil {
+		if entry.RemoteEntry.RemoteMtime > entry.Attributes.Mtime {
+			fileSize = maxUint64(fileSize, uint64(entry.RemoteEntry.RemoteSize))
+		}
+	}
+	return maxUint64(TotalSize(entry.Chunks), fileSize)
 }
 
 func ETag(entry *filer_pb.Entry) (etag string) {
@@ -245,19 +254,17 @@ func NonOverlappingVisibleIntervals(lookupFileIdFn wdclient.LookupFileIdFunction
 	if true {
 		return visibles2, err
 	}
-
-	sort.Slice(chunks, func(i, j int) bool {
-		if chunks[i].Mtime == chunks[j].Mtime {
-			filer_pb.EnsureFid(chunks[i])
-			filer_pb.EnsureFid(chunks[j])
-			if chunks[i].Fid == nil || chunks[j].Fid == nil {
+	slices.SortFunc(chunks, func(a, b *filer_pb.FileChunk) bool {
+		if a.Mtime == b.Mtime {
+			filer_pb.EnsureFid(a)
+			filer_pb.EnsureFid(b)
+			if a.Fid == nil || b.Fid == nil {
 				return true
 			}
-			return chunks[i].Fid.FileKey < chunks[j].Fid.FileKey
+			return a.Fid.FileKey < b.Fid.FileKey
 		}
-		return chunks[i].Mtime < chunks[j].Mtime // keep this to make tests run
+		return a.Mtime < b.Mtime
 	})
-
 	for _, chunk := range chunks {
 
 		// glog.V(0).Infof("merge [%d,%d)", chunk.Offset, chunk.Offset+int64(chunk.Size))

weed/filer/filechunks2_test.go

@@ -1,7 +1,7 @@
 package filer
 
 import (
-	"sort"
+	"golang.org/x/exp/slices"
 	"testing"
 
 	"github.com/chrislusf/seaweedfs/weed/glog"
@@ -34,11 +34,11 @@ func TestCompactFileChunksRealCase(t *testing.T) {
 }
 
 func printChunks(name string, chunks []*filer_pb.FileChunk) {
-	sort.Slice(chunks, func(i, j int) bool {
-		if chunks[i].Offset == chunks[j].Offset {
-			return chunks[i].Mtime < chunks[j].Mtime
+	slices.SortFunc(chunks, func(a, b *filer_pb.FileChunk) bool {
+		if a.Offset == b.Offset {
+			return a.Mtime < b.Mtime
 		}
-		return chunks[i].Offset < chunks[j].Offset
+		return a.Offset < b.Offset
 	})
 	for _, chunk := range chunks {
 		glog.V(0).Infof("%s chunk %s [%10d,%10d)", name, chunk.GetFileIdString(), chunk.Offset, chunk.Offset+int64(chunk.Size))

weed/filer/filechunks_read.go

@@ -2,7 +2,7 @@ package filer
 
 import (
 	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"sort"
+	"golang.org/x/exp/slices"
 )
 
 func readResolvedChunks(chunks []*filer_pb.FileChunk) (visibles []VisibleInterval) {
@@ -22,17 +22,14 @@ func readResolvedChunks(chunks []*filer_pb.FileChunk) (visibles []VisibleInterva
 			isStart: false,
 		})
 	}
-	sort.Slice(points, func(i, j int) bool {
-		if points[i].x != points[j].x {
-			return points[i].x < points[j].x
+	slices.SortFunc(points, func(a, b *Point) bool {
+		if a.x != b.x {
+			return a.x < b.x
 		}
-		if points[i].ts != points[j].ts {
-			return points[i].ts < points[j].ts
+		if a.ts != b.ts {
+			return a.ts < b.ts
 		}
-		if !points[i].isStart {
-			return true
-		}
-		return false
+		return !a.isStart
 	})
 
 	var prevX int64

weed/filer/filer.go

@@ -49,7 +49,7 @@ type Filer struct {
 	UniqueFileId        uint32
 }
 
-func NewFiler(masters []pb.ServerAddress, grpcDialOption grpc.DialOption,
+func NewFiler(masters map[string]pb.ServerAddress, grpcDialOption grpc.DialOption,
 	filerHost pb.ServerAddress, collection string, replication string, dataCenter string, notifyFn func()) *Filer {
 	f := &Filer{
 		MasterClient:        wdclient.NewMasterClient(grpcDialOption, cluster.FilerType, filerHost, dataCenter, masters),
@@ -151,7 +151,7 @@ func (f *Filer) RollbackTransaction(ctx context.Context) error {
 	return f.Store.RollbackTransaction(ctx)
 }
 
-func (f *Filer) CreateEntry(ctx context.Context, entry *Entry, o_excl bool, isFromOtherCluster bool, signatures []int32) error {
+func (f *Filer) CreateEntry(ctx context.Context, entry *Entry, o_excl bool, isFromOtherCluster bool, signatures []int32, skipCreateParentDir bool) error {
 
 	if string(entry.FullPath) == "/" {
 		return nil
@@ -169,9 +169,11 @@ func (f *Filer) CreateEntry(ctx context.Context, entry *Entry, o_excl bool, isFr
 
 	if oldEntry == nil {
 
-		dirParts := strings.Split(string(entry.FullPath), "/")
-		if err := f.ensureParentDirecotryEntry(ctx, entry, dirParts, len(dirParts)-1, isFromOtherCluster); err != nil {
-			return err
+		if !skipCreateParentDir {
+			dirParts := strings.Split(string(entry.FullPath), "/")
+			if err := f.ensureParentDirecotryEntry(ctx, entry, dirParts, len(dirParts)-1, isFromOtherCluster); err != nil {
+				return err
+			}
 		}
 
 		glog.V(4).Infof("InsertEntry %s: new entry: %v", entry.FullPath, entry.Name())

weed/filer/filer_delete_entry.go

@@ -9,8 +9,6 @@ import (
 	"github.com/chrislusf/seaweedfs/weed/util"
 )
 
-type HardLinkId []byte
-
 const (
 	MsgFailDelNonEmptyFolder = "fail to delete non-empty folder"
 )
@@ -127,7 +125,11 @@ func (f *Filer) doDeleteEntryMetaAndData(ctx context.Context, entry *Entry, shou
 
 	glog.V(3).Infof("deleting entry %v, delete chunks: %v", entry.FullPath, shouldDeleteChunks)
 
-	if storeDeletionErr := f.Store.DeleteOneEntry(ctx, entry); storeDeletionErr != nil {
+	if !entry.IsDirectory() && !shouldDeleteChunks {
+		if storeDeletionErr := f.Store.DeleteOneEntrySkipHardlink(ctx, entry.FullPath); storeDeletionErr != nil {
+			return fmt.Errorf("filer store delete skip hardlink: %v", storeDeletionErr)
+		}
+	} else if storeDeletionErr := f.Store.DeleteOneEntry(ctx, entry); storeDeletionErr != nil {
 		return fmt.Errorf("filer store delete: %v", storeDeletionErr)
 	}
 	if !entry.IsDirectory() {

weed/filer/filer_deletion.go

@@ -1,6 +1,7 @@
 package filer
 
 import (
+	"math"
 	"strings"
 	"time"
 
@@ -129,6 +130,12 @@ func (f *Filer) DeleteChunks(chunks []*filer_pb.FileChunk) {
 	}
 }
 
+func (f *Filer) DeleteChunksNotRecursive(chunks []*filer_pb.FileChunk) {
+	for _, chunk := range chunks {
+		f.fileIdDeletionQueue.EnQueue(chunk.GetFileIdString())
+	}
+}
+
 func (f *Filer) deleteChunksIfNotNew(oldEntry, newEntry *Entry) {
 
 	if oldEntry == nil {
@@ -141,14 +148,36 @@ func (f *Filer) deleteChunksIfNotNew(oldEntry, newEntry *Entry) {
 
 	var toDelete []*filer_pb.FileChunk
 	newChunkIds := make(map[string]bool)
-	for _, newChunk := range newEntry.Chunks {
+	newDataChunks, newManifestChunks, err := ResolveChunkManifest(f.MasterClient.GetLookupFileIdFunction(),
+		newEntry.Chunks, 0, math.MaxInt64)
+	if err != nil {
+		glog.Errorf("Failed to resolve new entry chunks when delete old entry chunks. new: %s, old: %s",
+			newEntry.Chunks, oldEntry.Chunks)
+		return
+	}
+	for _, newChunk := range newDataChunks {
+		newChunkIds[newChunk.GetFileIdString()] = true
+	}
+	for _, newChunk := range newManifestChunks {
 		newChunkIds[newChunk.GetFileIdString()] = true
 	}
 
-	for _, oldChunk := range oldEntry.Chunks {
+	oldDataChunks, oldManifestChunks, err := ResolveChunkManifest(f.MasterClient.GetLookupFileIdFunction(),
+		oldEntry.Chunks, 0, math.MaxInt64)
+	if err != nil {
+		glog.Errorf("Failed to resolve old entry chunks when delete old entry chunks. new: %s, old: %s",
+			newEntry.Chunks, oldEntry.Chunks)
+		return
+	}
+	for _, oldChunk := range oldDataChunks {
 		if _, found := newChunkIds[oldChunk.GetFileIdString()]; !found {
 			toDelete = append(toDelete, oldChunk)
 		}
 	}
-	f.DeleteChunks(toDelete)
+	for _, oldChunk := range oldManifestChunks {
+		if _, found := newChunkIds[oldChunk.GetFileIdString()]; !found {
+			toDelete = append(toDelete, oldChunk)
+		}
+	}
+	f.DeleteChunksNotRecursive(toDelete)
 }

weed/filer/filer_hardlink.go

@@ -0,0 +1,16 @@
+package filer
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+const (
+	HARD_LINK_MARKER = '\x01'
+)
+
+type HardLinkId []byte // 16 bytes + 1 marker byte
+
+func NewHardLinkId() HardLinkId {
+	bytes := append(util.RandomBytes(16), HARD_LINK_MARKER)
+	return bytes
+}

weed/filer/filer_notify_append.go

@@ -43,7 +43,7 @@ func (f *Filer) appendToFile(targetFile string, data []byte) error {
 	entry.Chunks = append(entry.Chunks, uploadResult.ToPbFileChunk(assignResult.Fid, offset))
 
 	// update the entry
-	err = f.CreateEntry(context.Background(), entry, false, false, nil)
+	err = f.CreateEntry(context.Background(), entry, false, false, nil, false)
 
 	return err
 }

weed/filer/filer_on_meta_event.go

@@ -22,12 +22,12 @@ func (f *Filer) onBucketEvents(event *filer_pb.SubscribeMetadataResponse) {
 		}
 	}
 	if f.DirBucketsPath == event.Directory {
-		if message.OldEntry == nil && message.NewEntry != nil {
+		if filer_pb.IsCreate(event) {
 			if message.NewEntry.IsDirectory {
 				f.Store.OnBucketCreation(message.NewEntry.Name)
 			}
 		}
-		if message.OldEntry != nil && message.NewEntry == nil {
+		if filer_pb.IsDelete(event) {
 			if message.OldEntry.IsDirectory {
 				f.Store.OnBucketDeletion(message.OldEntry.Name)
 			}

weed/filer/filerstore_hardlink.go

@@ -9,16 +9,20 @@ import (
 )
 
 func (fsw *FilerStoreWrapper) handleUpdateToHardLinks(ctx context.Context, entry *Entry) error {
-	if len(entry.HardLinkId) == 0 {
+
+	if entry.IsDirectory() {
 		return nil
 	}
-	// handle hard links
-	if err := fsw.setHardLink(ctx, entry); err != nil {
-		return fmt.Errorf("setHardLink %d: %v", entry.HardLinkId, err)
+
+	if len(entry.HardLinkId) > 0 {
+		// handle hard links
+		if err := fsw.setHardLink(ctx, entry); err != nil {
+			return fmt.Errorf("setHardLink %d: %v", entry.HardLinkId, err)
+		}
 	}
 
 	// check what is existing entry
-	glog.V(4).Infof("handleUpdateToHardLinks FindEntry %s", entry.FullPath)
+	// glog.V(4).Infof("handleUpdateToHardLinks FindEntry %s", entry.FullPath)
 	actualStore := fsw.getActualStore(entry.FullPath)
 	existingEntry, err := actualStore.FindEntry(ctx, entry.FullPath)
 	if err != nil && err != filer_pb.ErrNotFound {
@@ -46,6 +50,8 @@ func (fsw *FilerStoreWrapper) setHardLink(ctx context.Context, entry *Entry) err
 		return encodeErr
 	}
 
+	glog.V(4).Infof("setHardLink %v nlink:%d", entry.FullPath, entry.HardLinkCounter)
+
 	return fsw.KvPut(ctx, key, newBlob)
 }
 
@@ -55,7 +61,6 @@ func (fsw *FilerStoreWrapper) maybeReadHardLink(ctx context.Context, entry *Entr
 	}
 	key := entry.HardLinkId
 
-	glog.V(4).Infof("maybeReadHardLink KvGet %v", key)
 	value, err := fsw.KvGet(ctx, key)
 	if err != nil {
 		glog.Errorf("read %s hardlink %d: %v", entry.FullPath, entry.HardLinkId, err)
@@ -67,6 +72,8 @@ func (fsw *FilerStoreWrapper) maybeReadHardLink(ctx context.Context, entry *Entr
 		return err
 	}
 
+	glog.V(4).Infof("maybeReadHardLink %v nlink:%d", entry.FullPath, entry.HardLinkCounter)
+
 	return nil
 }
 

weed/filer/filerstore_wrapper.go

@@ -23,6 +23,7 @@ type VirtualFilerStore interface {
 	FilerStore
 	DeleteHardLink(ctx context.Context, hardLinkId HardLinkId) error
 	DeleteOneEntry(ctx context.Context, entry *Entry) error
+	DeleteOneEntrySkipHardlink(ctx context.Context, fullpath util.FullPath) error
 	AddPathSpecificStore(path string, storeId string, store FilerStore)
 	OnBucketCreation(bucket string)
 	OnBucketDeletion(bucket string)
@@ -127,7 +128,7 @@ func (fsw *FilerStoreWrapper) InsertEntry(ctx context.Context, entry *Entry) err
 		return err
 	}
 
-	glog.V(4).Infof("InsertEntry %s", entry.FullPath)
+	// glog.V(4).Infof("InsertEntry %s", entry.FullPath)
 	return actualStore.InsertEntry(ctx, entry)
 }
 
@@ -148,7 +149,7 @@ func (fsw *FilerStoreWrapper) UpdateEntry(ctx context.Context, entry *Entry) err
 		return err
 	}
 
-	glog.V(4).Infof("UpdateEntry %s", entry.FullPath)
+	// glog.V(4).Infof("UpdateEntry %s", entry.FullPath)
 	return actualStore.UpdateEntry(ctx, entry)
 }
 
@@ -192,7 +193,7 @@ func (fsw *FilerStoreWrapper) DeleteEntry(ctx context.Context, fp util.FullPath)
 		}
 	}
 
-	glog.V(4).Infof("DeleteEntry %s", fp)
+	// glog.V(4).Infof("DeleteEntry %s", fp)
 	return actualStore.DeleteEntry(ctx, fp)
 }
 
@@ -212,10 +213,22 @@ func (fsw *FilerStoreWrapper) DeleteOneEntry(ctx context.Context, existingEntry
 		}
 	}
 
-	glog.V(4).Infof("DeleteOneEntry %s", existingEntry.FullPath)
+	// glog.V(4).Infof("DeleteOneEntry %s", existingEntry.FullPath)
 	return actualStore.DeleteEntry(ctx, existingEntry.FullPath)
 }
 
+func (fsw *FilerStoreWrapper) DeleteOneEntrySkipHardlink(ctx context.Context, fullpath util.FullPath) (err error) {
+	actualStore := fsw.getActualStore(fullpath)
+	stats.FilerStoreCounter.WithLabelValues(actualStore.GetName(), "delete").Inc()
+	start := time.Now()
+	defer func() {
+		stats.FilerStoreHistogram.WithLabelValues(actualStore.GetName(), "delete").Observe(time.Since(start).Seconds())
+	}()
+
+	glog.V(4).Infof("DeleteOneEntrySkipHardlink %s", fullpath)
+	return actualStore.DeleteEntry(ctx, fullpath)
+}
+
 func (fsw *FilerStoreWrapper) DeleteFolderChildren(ctx context.Context, fp util.FullPath) (err error) {
 	actualStore := fsw.getActualStore(fp + "/")
 	stats.FilerStoreCounter.WithLabelValues(actualStore.GetName(), "deleteFolderChildren").Inc()
@@ -224,7 +237,7 @@ func (fsw *FilerStoreWrapper) DeleteFolderChildren(ctx context.Context, fp util.
 		stats.FilerStoreHistogram.WithLabelValues(actualStore.GetName(), "deleteFolderChildren").Observe(time.Since(start).Seconds())
 	}()
 
-	glog.V(4).Infof("DeleteFolderChildren %s", fp)
+	// glog.V(4).Infof("DeleteFolderChildren %s", fp)
 	return actualStore.DeleteFolderChildren(ctx, fp)
 }
 
@@ -236,7 +249,7 @@ func (fsw *FilerStoreWrapper) ListDirectoryEntries(ctx context.Context, dirPath
 		stats.FilerStoreHistogram.WithLabelValues(actualStore.GetName(), "list").Observe(time.Since(start).Seconds())
 	}()
 
-	glog.V(4).Infof("ListDirectoryEntries %s from %s limit %d", dirPath, startFileName, limit)
+	// glog.V(4).Infof("ListDirectoryEntries %s from %s limit %d", dirPath, startFileName, limit)
 	return actualStore.ListDirectoryEntries(ctx, dirPath, startFileName, includeStartFile, limit, func(entry *Entry) bool {
 		fsw.maybeReadHardLink(ctx, entry)
 		filer_pb.AfterEntryDeserialization(entry.Chunks)
@@ -254,7 +267,7 @@ func (fsw *FilerStoreWrapper) ListDirectoryPrefixedEntries(ctx context.Context,
 	if limit > math.MaxInt32-1 {
 		limit = math.MaxInt32 - 1
 	}
-	glog.V(4).Infof("ListDirectoryPrefixedEntries %s from %s prefix %s limit %d", dirPath, startFileName, prefix, limit)
+	// glog.V(4).Infof("ListDirectoryPrefixedEntries %s from %s prefix %s limit %d", dirPath, startFileName, prefix, limit)
 	adjustedEntryFunc := func(entry *Entry) bool {
 		fsw.maybeReadHardLink(ctx, entry)
 		filer_pb.AfterEntryDeserialization(entry.Chunks)
@@ -285,8 +298,10 @@ func (fsw *FilerStoreWrapper) prefixFilterEntries(ctx context.Context, dirPath u
 
 	count := int64(0)
 	for count < limit && len(notPrefixed) > 0 {
+		var isLastItemHasPrefix bool
 		for _, entry := range notPrefixed {
 			if strings.HasPrefix(entry.Name(), prefix) {
+				isLastItemHasPrefix = true
 				count++
 				if !eachEntryFunc(entry) {
 					return
@@ -294,9 +309,11 @@ func (fsw *FilerStoreWrapper) prefixFilterEntries(ctx context.Context, dirPath u
 				if count >= limit {
 					break
 				}
+			} else {
+				isLastItemHasPrefix = false
 			}
 		}
-		if count < limit {
+		if count < limit && isLastItemHasPrefix && len(notPrefixed) == int(limit) {
 			notPrefixed = notPrefixed[:0]
 			lastFileName, err = actualStore.ListDirectoryEntries(ctx, dirPath, lastFileName, false, limit, func(entry *Entry) bool {
 				notPrefixed = append(notPrefixed, entry)
@@ -305,6 +322,8 @@ func (fsw *FilerStoreWrapper) prefixFilterEntries(ctx context.Context, dirPath u
 			if err != nil {
 				return
 			}
+		} else {
+			break
 		}
 	}
 	return

weed/filer/leveldb/leveldb_store_test.go

@@ -13,8 +13,7 @@ import (
 
 func TestCreateAndFind(t *testing.T) {
 	testFiler := filer.NewFiler(nil, nil, "", "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_test")
-	defer os.RemoveAll(dir)
+	dir := t.TempDir()
 	store := &LevelDBStore{}
 	store.initialize(dir)
 	testFiler.SetStore(store)
@@ -32,7 +31,7 @@ func TestCreateAndFind(t *testing.T) {
 		},
 	}
 
-	if err := testFiler.CreateEntry(ctx, entry1, false, false, nil); err != nil {
+	if err := testFiler.CreateEntry(ctx, entry1, false, false, nil, false); err != nil {
 		t.Errorf("create entry %v: %v", entry1.FullPath, err)
 		return
 	}
@@ -67,8 +66,7 @@ func TestCreateAndFind(t *testing.T) {
 
 func TestEmptyRoot(t *testing.T) {
 	testFiler := filer.NewFiler(nil, nil, "", "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_test2")
-	defer os.RemoveAll(dir)
+	dir := t.TempDir()
 	store := &LevelDBStore{}
 	store.initialize(dir)
 	testFiler.SetStore(store)
@@ -90,8 +88,7 @@ func TestEmptyRoot(t *testing.T) {
 
 func BenchmarkInsertEntry(b *testing.B) {
 	testFiler := filer.NewFiler(nil, nil, "", "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_bench")
-	defer os.RemoveAll(dir)
+	dir := b.TempDir()
 	store := &LevelDBStore{}
 	store.initialize(dir)
 	testFiler.SetStore(store)

weed/filer/leveldb2/leveldb2_store_test.go

@@ -2,7 +2,6 @@ package leveldb
 
 import (
 	"context"
-	"os"
 	"testing"
 
 	"github.com/chrislusf/seaweedfs/weed/filer"
@@ -11,8 +10,7 @@ import (
 
 func TestCreateAndFind(t *testing.T) {
 	testFiler := filer.NewFiler(nil, nil, "", "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_test")
-	defer os.RemoveAll(dir)
+	dir := t.TempDir()
 	store := &LevelDB2Store{}
 	store.initialize(dir, 2)
 	testFiler.SetStore(store)
@@ -30,7 +28,7 @@ func TestCreateAndFind(t *testing.T) {
 		},
 	}
 
-	if err := testFiler.CreateEntry(ctx, entry1, false, false, nil); err != nil {
+	if err := testFiler.CreateEntry(ctx, entry1, false, false, nil, false); err != nil {
 		t.Errorf("create entry %v: %v", entry1.FullPath, err)
 		return
 	}
@@ -65,8 +63,7 @@ func TestCreateAndFind(t *testing.T) {
 
 func TestEmptyRoot(t *testing.T) {
 	testFiler := filer.NewFiler(nil, nil, "", "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_test2")
-	defer os.RemoveAll(dir)
+	dir := t.TempDir()
 	store := &LevelDB2Store{}
 	store.initialize(dir, 2)
 	testFiler.SetStore(store)

weed/filer/leveldb3/leveldb3_store.go

@@ -72,8 +72,8 @@ func (store *LevelDB3Store) loadDB(name string) (*leveldb.DB, error) {
 	}
 	if name != DEFAULT {
 		opts = &opt.Options{
-			BlockCacheCapacity: 4 * 1024 * 1024, // default value is 8MiB
-			WriteBuffer:        2 * 1024 * 1024, // default value is 4MiB
+			BlockCacheCapacity: 16 * 1024 * 1024, // default value is 8MiB
+			WriteBuffer:        8 * 1024 * 1024,  // default value is 4MiB
 			Filter:             bloom,
 		}
 	}

weed/filer/leveldb3/leveldb3_store_test.go

@@ -2,7 +2,6 @@ package leveldb
 
 import (
 	"context"
-	"os"
 	"testing"
 
 	"github.com/chrislusf/seaweedfs/weed/filer"
@@ -11,8 +10,7 @@ import (
 
 func TestCreateAndFind(t *testing.T) {
 	testFiler := filer.NewFiler(nil, nil, "", "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_test")
-	defer os.RemoveAll(dir)
+	dir := t.TempDir()
 	store := &LevelDB3Store{}
 	store.initialize(dir)
 	testFiler.SetStore(store)
@@ -30,7 +28,7 @@ func TestCreateAndFind(t *testing.T) {
 		},
 	}
 
-	if err := testFiler.CreateEntry(ctx, entry1, false, false, nil); err != nil {
+	if err := testFiler.CreateEntry(ctx, entry1, false, false, nil, false); err != nil {
 		t.Errorf("create entry %v: %v", entry1.FullPath, err)
 		return
 	}
@@ -65,8 +63,7 @@ func TestCreateAndFind(t *testing.T) {
 
 func TestEmptyRoot(t *testing.T) {
 	testFiler := filer.NewFiler(nil, nil, "", "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_test2")
-	defer os.RemoveAll(dir)
+	dir := t.TempDir()
 	store := &LevelDB3Store{}
 	store.initialize(dir)
 	testFiler.SetStore(store)

weed/filer/meta_aggregator.go

@@ -76,9 +76,6 @@ func (ma *MetaAggregator) setActive(address pb.ServerAddress, isActive bool) (no
 		}
 	} else {
 		if _, found := ma.peerStatues[address]; found {
-			ma.peerStatues[address] -= 1
-		}
-		if ma.peerStatues[address] <= 0 {
 			delete(ma.peerStatues, address)
 		}
 	}

weed/filer/mongodb/mongodb_store.go

@@ -159,7 +159,7 @@ func (store *MongodbStore) DeleteEntry(ctx context.Context, fullpath util.FullPa
 	dir, name := fullpath.DirAndName()
 
 	where := bson.M{"directory": dir, "name": name}
-	_, err := store.connect.Database(store.database).Collection(store.collectionName).DeleteOne(ctx, where)
+	_, err := store.connect.Database(store.database).Collection(store.collectionName).DeleteMany(ctx, where)
 	if err != nil {
 		return fmt.Errorf("delete %s : %v", fullpath, err)
 	}
@@ -199,9 +199,9 @@ func (store *MongodbStore) ListDirectoryEntries(ctx context.Context, dirPath uti
 
 	for cur.Next(ctx) {
 		var data Model
-		err := cur.Decode(&data)
-		if err != nil && err != mongo.ErrNoDocuments {
-			return lastFileName, err
+		err = cur.Decode(&data)
+		if err != nil {
+			break
 		}
 
 		entry := &filer.Entry{

weed/filer/mysql2/mysql2_store.go

@@ -4,6 +4,7 @@ import (
 	"context"
 	"database/sql"
 	"fmt"
+	"strings"
 	"time"
 
 	"github.com/chrislusf/seaweedfs/weed/filer"
@@ -82,7 +83,7 @@ func (store *MysqlStore2) initialize(createTable, upsertQuery string, enableUpse
 		return fmt.Errorf("connect to %s error:%v", sqlUrl, err)
 	}
 
-	if err = store.CreateTable(context.Background(), abstract_sql.DEFAULT_TABLE); err != nil {
+	if err = store.CreateTable(context.Background(), abstract_sql.DEFAULT_TABLE); err != nil && !strings.Contains(err.Error(), "table already exist") {
 		return fmt.Errorf("init table %s: %v", abstract_sql.DEFAULT_TABLE, err)
 	}
 

weed/filer/read_write.go

@@ -62,6 +62,7 @@ func SaveInsideFiler(client filer_pb.SeaweedFilerClient, dir, name string, conte
 				},
 				Content: content,
 			},
+			SkipCheckParentDirectory: true,
 		})
 	} else if err == nil {
 		entry := resp.Entry

weed/filer/reader_at.go

@@ -12,21 +12,16 @@ import (
 	"github.com/chrislusf/seaweedfs/weed/util"
 	"github.com/chrislusf/seaweedfs/weed/util/chunk_cache"
 	"github.com/chrislusf/seaweedfs/weed/wdclient"
-	"github.com/golang/groupcache/singleflight"
 )
 
 type ChunkReadAt struct {
-	masterClient *wdclient.MasterClient
-	chunkViews   []*ChunkView
-	lookupFileId wdclient.LookupFileIdFunctionType
-	readerLock   sync.Mutex
-	fileSize     int64
-
-	fetchGroup      singleflight.Group
-	chunkCache      chunk_cache.ChunkCache
-	lastChunkFileId string
-	lastChunkData   []byte
-	readerPattern   *ReaderPattern
+	masterClient  *wdclient.MasterClient
+	chunkViews    []*ChunkView
+	readerLock    sync.Mutex
+	fileSize      int64
+	readerCache   *ReaderCache
+	readerPattern *ReaderPattern
+	lastChunkFid  string
 }
 
 var _ = io.ReaderAt(&ChunkReadAt{})
@@ -90,16 +85,14 @@ func NewChunkReaderAtFromClient(lookupFn wdclient.LookupFileIdFunctionType, chun
 
 	return &ChunkReadAt{
 		chunkViews:    chunkViews,
-		lookupFileId:  lookupFn,
-		chunkCache:    chunkCache,
 		fileSize:      fileSize,
+		readerCache:   newReaderCache(32, chunkCache, lookupFn),
 		readerPattern: NewReaderPattern(),
 	}
 }
 
 func (c *ChunkReadAt) Close() error {
-	c.lastChunkData = nil
-	c.lastChunkFileId = ""
+	c.readerCache.destroy()
 	return nil
 }
 
@@ -117,15 +110,13 @@ func (c *ChunkReadAt) ReadAt(p []byte, offset int64) (n int, err error) {
 func (c *ChunkReadAt) doReadAt(p []byte, offset int64) (n int, err error) {
 
 	startOffset, remaining := offset, int64(len(p))
-	var nextChunk *ChunkView
+	var nextChunks []*ChunkView
 	for i, chunk := range c.chunkViews {
 		if remaining <= 0 {
 			break
 		}
 		if i+1 < len(c.chunkViews) {
-			nextChunk = c.chunkViews[i+1]
-		} else {
-			nextChunk = nil
+			nextChunks = c.chunkViews[i+1:]
 		}
 		if startOffset < chunk.LogicOffset {
 			gap := int(chunk.LogicOffset - startOffset)
@@ -142,16 +133,13 @@ func (c *ChunkReadAt) doReadAt(p []byte, offset int64) (n int, err error) {
 			continue
 		}
 		// glog.V(4).Infof("read [%d,%d), %d/%d chunk %s [%d,%d)", chunkStart, chunkStop, i, len(c.chunkViews), chunk.FileId, chunk.LogicOffset-chunk.Offset, chunk.LogicOffset-chunk.Offset+int64(chunk.Size))
-		var buffer []byte
 		bufferOffset := chunkStart - chunk.LogicOffset + chunk.Offset
-		bufferLength := chunkStop - chunkStart
-		buffer, err = c.readChunkSlice(chunk, nextChunk, uint64(bufferOffset), uint64(bufferLength))
+		copied, err := c.readChunkSliceAt(p[startOffset-offset:chunkStop-chunkStart+startOffset-offset], chunk, nextChunks, uint64(bufferOffset))
 		if err != nil {
 			glog.Errorf("fetching chunk %+v: %v\n", chunk, err)
-			return
+			return copied, err
 		}
 
-		copied := copy(p[startOffset-offset:chunkStop-chunkStart+startOffset-offset], buffer)
 		n += copied
 		startOffset, remaining = startOffset+int64(copied), remaining-int64(copied)
 	}
@@ -173,104 +161,25 @@ func (c *ChunkReadAt) doReadAt(p []byte, offset int64) (n int, err error) {
 
 }
 
-func (c *ChunkReadAt) readChunkSlice(chunkView *ChunkView, nextChunkViews *ChunkView, offset, length uint64) ([]byte, error) {
+func (c *ChunkReadAt) readChunkSliceAt(buffer []byte, chunkView *ChunkView, nextChunkViews []*ChunkView, offset uint64) (n int, err error) {
 
-	var chunkSlice []byte
-	if chunkView.LogicOffset == 0 {
-		chunkSlice = c.chunkCache.GetChunkSlice(chunkView.FileId, offset, length)
-	}
-	if len(chunkSlice) > 0 {
-		return chunkSlice, nil
-	}
-	if c.lookupFileId == nil {
-		return nil, nil
-	}
 	if c.readerPattern.IsRandomMode() {
-		return c.doFetchRangeChunkData(chunkView, offset, length)
-	}
-	chunkData, err := c.readFromWholeChunkData(chunkView, nextChunkViews)
-	if err != nil {
-		return nil, err
-	}
-	wanted := min(int64(length), int64(len(chunkData))-int64(offset))
-	return chunkData[offset : int64(offset)+wanted], nil
-}
-
-func (c *ChunkReadAt) readFromWholeChunkData(chunkView *ChunkView, nextChunkViews ...*ChunkView) (chunkData []byte, err error) {
-
-	if c.lastChunkFileId == chunkView.FileId {
-		return c.lastChunkData, nil
+		return fetchChunkRange(buffer, c.readerCache.lookupFileIdFn, chunkView.FileId, chunkView.CipherKey, chunkView.IsGzipped, int64(offset))
 	}
 
-	v, doErr := c.readOneWholeChunk(chunkView)
-
-	if doErr != nil {
-		return nil, doErr
-	}
-
-	chunkData = v.([]byte)
-
-	c.lastChunkData = chunkData
-	c.lastChunkFileId = chunkView.FileId
-
-	for _, nextChunkView := range nextChunkViews {
-		if c.chunkCache != nil && nextChunkView != nil {
-			go c.readOneWholeChunk(nextChunkView)
+	n, err = c.readerCache.ReadChunkAt(buffer, chunkView.FileId, chunkView.CipherKey, chunkView.IsGzipped, int64(offset), int(chunkView.ChunkSize), chunkView.LogicOffset == 0)
+	if c.lastChunkFid != chunkView.FileId {
+		if chunkView.Offset == 0 { // start of a new chunk
+			if c.lastChunkFid != "" {
+				c.readerCache.UnCache(c.lastChunkFid)
+				c.readerCache.MaybeCache(nextChunkViews)
+			} else {
+				if len(nextChunkViews) >= 1 {
+					c.readerCache.MaybeCache(nextChunkViews[:1]) // just read the next chunk if at the very beginning
+				}
+			}
 		}
 	}
-
+	c.lastChunkFid = chunkView.FileId
 	return
 }
-
-func (c *ChunkReadAt) readOneWholeChunk(chunkView *ChunkView) (interface{}, error) {
-
-	var err error
-
-	return c.fetchGroup.Do(chunkView.FileId, func() (interface{}, error) {
-
-		glog.V(4).Infof("readFromWholeChunkData %s offset %d [%d,%d) size at least %d", chunkView.FileId, chunkView.Offset, chunkView.LogicOffset, chunkView.LogicOffset+int64(chunkView.Size), chunkView.ChunkSize)
-
-		var data []byte
-		if chunkView.LogicOffset == 0 {
-			data = c.chunkCache.GetChunk(chunkView.FileId, chunkView.ChunkSize)
-		}
-		if data != nil {
-			glog.V(4).Infof("cache hit %s [%d,%d)", chunkView.FileId, chunkView.LogicOffset-chunkView.Offset, chunkView.LogicOffset-chunkView.Offset+int64(len(data)))
-		} else {
-			var err error
-			data, err = c.doFetchFullChunkData(chunkView)
-			if err != nil {
-				return data, err
-			}
-			if chunkView.LogicOffset == 0 {
-				// only cache the first chunk
-				c.chunkCache.SetChunk(chunkView.FileId, data)
-			}
-		}
-		return data, err
-	})
-}
-
-func (c *ChunkReadAt) doFetchFullChunkData(chunkView *ChunkView) ([]byte, error) {
-
-	glog.V(4).Infof("+ doFetchFullChunkData %s", chunkView.FileId)
-
-	data, err := fetchChunk(c.lookupFileId, chunkView.FileId, chunkView.CipherKey, chunkView.IsGzipped)
-
-	glog.V(4).Infof("- doFetchFullChunkData %s", chunkView.FileId)
-
-	return data, err
-
-}
-
-func (c *ChunkReadAt) doFetchRangeChunkData(chunkView *ChunkView, offset, length uint64) ([]byte, error) {
-
-	glog.V(4).Infof("+ doFetchFullChunkData %s", chunkView.FileId)
-
-	data, err := fetchChunkRange(c.lookupFileId, chunkView.FileId, chunkView.CipherKey, chunkView.IsGzipped, int64(offset), int(length))
-
-	glog.V(4).Infof("- doFetchFullChunkData %s", chunkView.FileId)
-
-	return data, err
-
-}

weed/filer/reader_at_test.go

@@ -21,8 +21,12 @@ func (m *mockChunkCache) GetChunk(fileId string, minSize uint64) (data []byte) {
 	return data
 }
 
-func (m *mockChunkCache) GetChunkSlice(fileId string, offset, length uint64) []byte {
-	return m.GetChunk(fileId, length)
+func (m *mockChunkCache) ReadChunkAt(data []byte, fileId string, offset uint64) (n int, err error) {
+	x, _ := strconv.Atoi(fileId)
+	for i := 0; i < len(data); i++ {
+		data[i] = byte(x)
+	}
+	return len(data), nil
 }
 
 func (m *mockChunkCache) SetChunk(fileId string, data []byte) {
@@ -65,10 +69,9 @@ func TestReaderAt(t *testing.T) {
 
 	readerAt := &ChunkReadAt{
 		chunkViews:    ViewFromVisibleIntervals(visibles, 0, math.MaxInt64),
-		lookupFileId:  nil,
 		readerLock:    sync.Mutex{},
 		fileSize:      10,
-		chunkCache:    &mockChunkCache{},
+		readerCache:   newReaderCache(3, &mockChunkCache{}, nil),
 		readerPattern: NewReaderPattern(),
 	}
 
@@ -81,7 +84,7 @@ func TestReaderAt(t *testing.T) {
 
 func testReadAt(t *testing.T, readerAt *ChunkReadAt, offset int64, size int, expected int, expectedErr error) {
 	data := make([]byte, size)
-	n, err := readerAt.ReadAt(data, offset)
+	n, err := readerAt.doReadAt(data, offset)
 
 	for _, d := range data {
 		fmt.Printf("%x", d)
@@ -116,10 +119,9 @@ func TestReaderAt0(t *testing.T) {
 
 	readerAt := &ChunkReadAt{
 		chunkViews:    ViewFromVisibleIntervals(visibles, 0, math.MaxInt64),
-		lookupFileId:  nil,
 		readerLock:    sync.Mutex{},
 		fileSize:      10,
-		chunkCache:    &mockChunkCache{},
+		readerCache:   newReaderCache(3, &mockChunkCache{}, nil),
 		readerPattern: NewReaderPattern(),
 	}
 
@@ -145,10 +147,9 @@ func TestReaderAt1(t *testing.T) {
 
 	readerAt := &ChunkReadAt{
 		chunkViews:    ViewFromVisibleIntervals(visibles, 0, math.MaxInt64),
-		lookupFileId:  nil,
 		readerLock:    sync.Mutex{},
 		fileSize:      20,
-		chunkCache:    &mockChunkCache{},
+		readerCache:   newReaderCache(3, &mockChunkCache{}, nil),
 		readerPattern: NewReaderPattern(),
 	}
 

weed/filer/reader_cache.go

@@ -0,0 +1,192 @@
+package filer
+
+import (
+	"fmt"
+	"github.com/chrislusf/seaweedfs/weed/util/chunk_cache"
+	"github.com/chrislusf/seaweedfs/weed/util/mem"
+	"github.com/chrislusf/seaweedfs/weed/wdclient"
+	"sync"
+	"time"
+)
+
+type ReaderCache struct {
+	chunkCache     chunk_cache.ChunkCache
+	lookupFileIdFn wdclient.LookupFileIdFunctionType
+	sync.Mutex
+	downloaders map[string]*SingleChunkCacher
+	limit       int
+}
+
+type SingleChunkCacher struct {
+	sync.RWMutex
+	parent        *ReaderCache
+	chunkFileId   string
+	data          []byte
+	err           error
+	cipherKey     []byte
+	isGzipped     bool
+	chunkSize     int
+	shouldCache   bool
+	wg            sync.WaitGroup
+	completedTime time.Time
+}
+
+func newReaderCache(limit int, chunkCache chunk_cache.ChunkCache, lookupFileIdFn wdclient.LookupFileIdFunctionType) *ReaderCache {
+	return &ReaderCache{
+		limit:          limit,
+		chunkCache:     chunkCache,
+		lookupFileIdFn: lookupFileIdFn,
+		downloaders:    make(map[string]*SingleChunkCacher),
+	}
+}
+
+func (rc *ReaderCache) MaybeCache(chunkViews []*ChunkView) {
+	if rc.lookupFileIdFn == nil {
+		return
+	}
+
+	rc.Lock()
+	defer rc.Unlock()
+
+	for _, chunkView := range chunkViews {
+		if _, found := rc.downloaders[chunkView.FileId]; found {
+			continue
+		}
+
+		if len(rc.downloaders) >= rc.limit {
+			// if still no slots, return
+			return
+		}
+
+		// glog.V(4).Infof("prefetch %s offset %d", chunkView.FileId, chunkView.LogicOffset)
+		// cache this chunk if not yet
+		cacher := newSingleChunkCacher(rc, chunkView.FileId, chunkView.CipherKey, chunkView.IsGzipped, int(chunkView.ChunkSize), false)
+		cacher.wg.Add(1)
+		go cacher.startCaching()
+		cacher.wg.Wait()
+		rc.downloaders[chunkView.FileId] = cacher
+
+	}
+
+	return
+}
+
+func (rc *ReaderCache) ReadChunkAt(buffer []byte, fileId string, cipherKey []byte, isGzipped bool, offset int64, chunkSize int, shouldCache bool) (int, error) {
+	rc.Lock()
+	defer rc.Unlock()
+	if cacher, found := rc.downloaders[fileId]; found {
+		return cacher.readChunkAt(buffer, offset)
+	}
+	if shouldCache || rc.lookupFileIdFn == nil {
+		n, err := rc.chunkCache.ReadChunkAt(buffer, fileId, uint64(offset))
+		if n > 0 {
+			return n, err
+		}
+	}
+
+	if len(rc.downloaders) >= rc.limit {
+		oldestFid, oldestTime := "", time.Now()
+		for fid, downloader := range rc.downloaders {
+			if !downloader.completedTime.IsZero() {
+				if downloader.completedTime.Before(oldestTime) {
+					oldestFid, oldestTime = fid, downloader.completedTime
+				}
+			}
+		}
+		if oldestFid != "" {
+			oldDownloader := rc.downloaders[oldestFid]
+			delete(rc.downloaders, oldestFid)
+			oldDownloader.destroy()
+		}
+	}
+
+	// glog.V(4).Infof("cache1 %s", fileId)
+
+	cacher := newSingleChunkCacher(rc, fileId, cipherKey, isGzipped, chunkSize, shouldCache)
+	cacher.wg.Add(1)
+	go cacher.startCaching()
+	cacher.wg.Wait()
+	rc.downloaders[fileId] = cacher
+
+	return cacher.readChunkAt(buffer, offset)
+}
+
+func (rc *ReaderCache) UnCache(fileId string) {
+	rc.Lock()
+	defer rc.Unlock()
+	// glog.V(4).Infof("uncache %s", fileId)
+	if downloader, found := rc.downloaders[fileId]; found {
+		downloader.destroy()
+		delete(rc.downloaders, fileId)
+	}
+}
+
+func (rc *ReaderCache) destroy() {
+	rc.Lock()
+	defer rc.Unlock()
+
+	for _, downloader := range rc.downloaders {
+		downloader.destroy()
+	}
+
+}
+
+func newSingleChunkCacher(parent *ReaderCache, fileId string, cipherKey []byte, isGzipped bool, chunkSize int, shouldCache bool) *SingleChunkCacher {
+	t := &SingleChunkCacher{
+		parent:      parent,
+		chunkFileId: fileId,
+		cipherKey:   cipherKey,
+		isGzipped:   isGzipped,
+		chunkSize:   chunkSize,
+		shouldCache: shouldCache,
+	}
+	return t
+}
+
+func (s *SingleChunkCacher) startCaching() {
+	s.Lock()
+	defer s.Unlock()
+
+	s.wg.Done() // means this has been started
+
+	urlStrings, err := s.parent.lookupFileIdFn(s.chunkFileId)
+	if err != nil {
+		s.err = fmt.Errorf("operation LookupFileId %s failed, err: %v", s.chunkFileId, err)
+		return
+	}
+
+	s.data = mem.Allocate(s.chunkSize)
+
+	_, s.err = retriedFetchChunkData(s.data, urlStrings, s.cipherKey, s.isGzipped, true, 0)
+	if s.err != nil {
+		mem.Free(s.data)
+		s.data = nil
+		return
+	}
+
+	s.completedTime = time.Now()
+	if s.shouldCache {
+		s.parent.chunkCache.SetChunk(s.chunkFileId, s.data)
+	}
+
+	return
+}
+
+func (s *SingleChunkCacher) destroy() {
+	if s.data != nil {
+		mem.Free(s.data)
+		s.data = nil
+	}
+}
+
+func (s *SingleChunkCacher) readChunkAt(buf []byte, offset int64) (int, error) {
+	s.RLock()
+	defer s.RUnlock()
+
+	if s.err != nil {
+		return 0, s.err
+	}
+
+	return copy(buf, s.data[offset:]), nil
+
+}

weed/filer/redis/universal_redis_store.go

@@ -3,7 +3,7 @@ package redis
 import (
 	"context"
 	"fmt"
-	"sort"
+	"golang.org/x/exp/slices"
 	"strings"
 	"time"
 
@@ -157,8 +157,8 @@ func (store *UniversalRedisStore) ListDirectoryEntries(ctx context.Context, dirP
 	}
 
 	// sort
-	sort.Slice(members, func(i, j int) bool {
-		return strings.Compare(members[i], members[j]) < 0
+	slices.SortFunc(members, func(a, b string) bool {
+		return strings.Compare(a, b) < 0
 	})
 
 	// limit

weed/filer/redis_lua/redis_cluster_store.go

@@ -0,0 +1,44 @@
+package redis_lua
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/go-redis/redis/v8"
+)
+
+func init() {
+	filer.Stores = append(filer.Stores, &RedisLuaClusterStore{})
+}
+
+type RedisLuaClusterStore struct {
+	UniversalRedisLuaStore
+}
+
+func (store *RedisLuaClusterStore) GetName() string {
+	return "redis_lua_cluster"
+}
+
+func (store *RedisLuaClusterStore) Initialize(configuration util.Configuration, prefix string) (err error) {
+
+	configuration.SetDefault(prefix+"useReadOnly", false)
+	configuration.SetDefault(prefix+"routeByLatency", false)
+
+	return store.initialize(
+		configuration.GetStringSlice(prefix+"addresses"),
+		configuration.GetString(prefix+"password"),
+		configuration.GetBool(prefix+"useReadOnly"),
+		configuration.GetBool(prefix+"routeByLatency"),
+		configuration.GetStringSlice(prefix+"superLargeDirectories"),
+	)
+}
+
+func (store *RedisLuaClusterStore) initialize(addresses []string, password string, readOnly, routeByLatency bool, superLargeDirectories []string) (err error) {
+	store.Client = redis.NewClusterClient(&redis.ClusterOptions{
+		Addrs:          addresses,
+		Password:       password,
+		ReadOnly:       readOnly,
+		RouteByLatency: routeByLatency,
+	})
+	store.loadSuperLargeDirectories(superLargeDirectories)
+	return
+}

weed/filer/redis_lua/redis_sentinel_store.go

@@ -0,0 +1,45 @@
+package redis_lua
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/go-redis/redis/v8"
+	"time"
+)
+
+func init() {
+	filer.Stores = append(filer.Stores, &RedisLuaSentinelStore{})
+}
+
+type RedisLuaSentinelStore struct {
+	UniversalRedisLuaStore
+}
+
+func (store *RedisLuaSentinelStore) GetName() string {
+	return "redis_lua_sentinel"
+}
+
+func (store *RedisLuaSentinelStore) Initialize(configuration util.Configuration, prefix string) (err error) {
+	return store.initialize(
+		configuration.GetStringSlice(prefix+"addresses"),
+		configuration.GetString(prefix+"masterName"),
+		configuration.GetString(prefix+"username"),
+		configuration.GetString(prefix+"password"),
+		configuration.GetInt(prefix+"database"),
+	)
+}
+
+func (store *RedisLuaSentinelStore) initialize(addresses []string, masterName string, username string, password string, database int) (err error) {
+	store.Client = redis.NewFailoverClient(&redis.FailoverOptions{
+		MasterName:      masterName,
+		SentinelAddrs:   addresses,
+		Username:        username,
+		Password:        password,
+		DB:              database,
+		MinRetryBackoff: time.Millisecond * 100,
+		MaxRetryBackoff: time.Minute * 1,
+		ReadTimeout:     time.Second * 30,
+		WriteTimeout:    time.Second * 5,
+	})
+	return
+}

weed/filer/redis_lua/redis_store.go

@@ -0,0 +1,38 @@
+package redis_lua
+
+import (
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/util"
+	"github.com/go-redis/redis/v8"
+)
+
+func init() {
+	filer.Stores = append(filer.Stores, &RedisLuaStore{})
+}
+
+type RedisLuaStore struct {
+	UniversalRedisLuaStore
+}
+
+func (store *RedisLuaStore) GetName() string {
+	return "redis_lua"
+}
+
+func (store *RedisLuaStore) Initialize(configuration util.Configuration, prefix string) (err error) {
+	return store.initialize(
+		configuration.GetString(prefix+"address"),
+		configuration.GetString(prefix+"password"),
+		configuration.GetInt(prefix+"database"),
+		configuration.GetStringSlice(prefix+"superLargeDirectories"),
+	)
+}
+
+func (store *RedisLuaStore) initialize(hostPort string, password string, database int, superLargeDirectories []string) (err error) {
+	store.Client = redis.NewClient(&redis.Options{
+		Addr:     hostPort,
+		Password: password,
+		DB:       database,
+	})
+	store.loadSuperLargeDirectories(superLargeDirectories)
+	return
+}

weed/filer/redis_lua/stored_procedure/delete_entry.lua

@@ -0,0 +1,19 @@
+-- KEYS[1]: full path of entry
+local fullpath = KEYS[1]
+-- KEYS[2]: full path of entry
+local fullpath_list_key = KEYS[2]
+-- KEYS[3]: dir of the entry
+local dir_list_key = KEYS[3]
+
+-- ARGV[1]: isSuperLargeDirectory
+local isSuperLargeDirectory = ARGV[1] == "1"
+-- ARGV[2]: name of the entry
+local name = ARGV[2]
+
+redis.call("DEL", fullpath, fullpath_list_key)
+
+if not isSuperLargeDirectory and name ~= "" then
+    redis.call("ZREM", dir_list_key, name)
+end
+
+return 0

weed/filer/redis_lua/stored_procedure/delete_folder_children.lua

@@ -0,0 +1,15 @@
+-- KEYS[1]: full path of entry
+local fullpath = KEYS[1]
+
+if fullpath ~= "" and string.sub(fullpath, -1) == "/" then
+    fullpath = string.sub(fullpath, 0, -2)
+end
+
+local files = redis.call("ZRANGE", fullpath .. "\0", "0", "-1")
+
+for _, name in ipairs(files) do
+    local file_path = fullpath .. "/" .. name
+    redis.call("DEL", file_path, file_path .. "\0")
+end
+
+return 0

weed/filer/redis_lua/stored_procedure/init.go

@@ -0,0 +1,24 @@
+package stored_procedure
+
+import (
+	_ "embed"
+	"github.com/go-redis/redis/v8"
+)
+
+func init() {
+	InsertEntryScript = redis.NewScript(insertEntry)
+	DeleteEntryScript = redis.NewScript(deleteEntry)
+	DeleteFolderChildrenScript = redis.NewScript(deleteFolderChildren)
+}
+
+//go:embed insert_entry.lua
+var insertEntry string
+var InsertEntryScript *redis.Script
+
+//go:embed delete_entry.lua
+var deleteEntry string
+var DeleteEntryScript *redis.Script
+
+//go:embed delete_folder_children.lua
+var deleteFolderChildren string
+var DeleteFolderChildrenScript *redis.Script

weed/filer/redis_lua/stored_procedure/insert_entry.lua

@@ -0,0 +1,27 @@
+-- KEYS[1]: full path of entry
+local full_path = KEYS[1]
+-- KEYS[2]: dir of the entry
+local dir_list_key = KEYS[2]
+
+-- ARGV[1]: content of the entry
+local entry = ARGV[1]
+-- ARGV[2]: TTL of the entry
+local ttlSec = tonumber(ARGV[2])
+-- ARGV[3]: isSuperLargeDirectory
+local isSuperLargeDirectory = ARGV[3] == "1"
+-- ARGV[4]: zscore of the entry in zset
+local zscore = tonumber(ARGV[4])
+-- ARGV[5]: name of the entry
+local name = ARGV[5]
+
+if ttlSec > 0 then
+    redis.call("SET", full_path, entry, "EX", ttlSec)
+else
+    redis.call("SET", full_path, entry)
+end
+
+if not isSuperLargeDirectory and name ~= "" then
+    redis.call("ZADD", dir_list_key, "NX", zscore, name)
+end
+
+return 0

weed/filer/redis_lua/universal_redis_store.go

@@ -0,0 +1,191 @@
+package redis_lua
+
+import (
+	"context"
+	"fmt"
+	"time"
+
+	"github.com/go-redis/redis/v8"
+
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/chrislusf/seaweedfs/weed/filer/redis_lua/stored_procedure"
+	"github.com/chrislusf/seaweedfs/weed/glog"
+	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
+	"github.com/chrislusf/seaweedfs/weed/util"
+)
+
+const (
+	DIR_LIST_MARKER = "\x00"
+)
+
+type UniversalRedisLuaStore struct {
+	Client                  redis.UniversalClient
+	superLargeDirectoryHash map[string]bool
+}
+
+func (store *UniversalRedisLuaStore) isSuperLargeDirectory(dir string) (isSuperLargeDirectory bool) {
+	_, isSuperLargeDirectory = store.superLargeDirectoryHash[dir]
+	return
+}
+
+func (store *UniversalRedisLuaStore) loadSuperLargeDirectories(superLargeDirectories []string) {
+	// set directory hash
+	store.superLargeDirectoryHash = make(map[string]bool)
+	for _, dir := range superLargeDirectories {
+		store.superLargeDirectoryHash[dir] = true
+	}
+}
+
+func (store *UniversalRedisLuaStore) BeginTransaction(ctx context.Context) (context.Context, error) {
+	return ctx, nil
+}
+func (store *UniversalRedisLuaStore) CommitTransaction(ctx context.Context) error {
+	return nil
+}
+func (store *UniversalRedisLuaStore) RollbackTransaction(ctx context.Context) error {
+	return nil
+}
+
+func (store *UniversalRedisLuaStore) InsertEntry(ctx context.Context, entry *filer.Entry) (err error) {
+
+	value, err := entry.EncodeAttributesAndChunks()
+	if err != nil {
+		return fmt.Errorf("encoding %s %+v: %v", entry.FullPath, entry.Attr, err)
+	}
+
+	if len(entry.Chunks) > 50 {
+		value = util.MaybeGzipData(value)
+	}
+
+	dir, name := entry.FullPath.DirAndName()
+
+	err = stored_procedure.InsertEntryScript.Run(ctx, store.Client,
+		[]string{string(entry.FullPath), genDirectoryListKey(dir)},
+		value, entry.TtlSec,
+		store.isSuperLargeDirectory(dir), 0, name).Err()
+
+	if err != nil {
+		return fmt.Errorf("persisting %s : %v", entry.FullPath, err)
+	}
+
+	return nil
+}
+
+func (store *UniversalRedisLuaStore) UpdateEntry(ctx context.Context, entry *filer.Entry) (err error) {
+
+	return store.InsertEntry(ctx, entry)
+}
+
+func (store *UniversalRedisLuaStore) FindEntry(ctx context.Context, fullpath util.FullPath) (entry *filer.Entry, err error) {
+
+	data, err := store.Client.Get(ctx, string(fullpath)).Result()
+	if err == redis.Nil {
+		return nil, filer_pb.ErrNotFound
+	}
+
+	if err != nil {
+		return nil, fmt.Errorf("get %s : %v", fullpath, err)
+	}
+
+	entry = &filer.Entry{
+		FullPath: fullpath,
+	}
+	err = entry.DecodeAttributesAndChunks(util.MaybeDecompressData([]byte(data)))
+	if err != nil {
+		return entry, fmt.Errorf("decode %s : %v", entry.FullPath, err)
+	}
+
+	return entry, nil
+}
+
+func (store *UniversalRedisLuaStore) DeleteEntry(ctx context.Context, fullpath util.FullPath) (err error) {
+
+	dir, name := fullpath.DirAndName()
+
+	err = stored_procedure.DeleteEntryScript.Run(ctx, store.Client,
+		[]string{string(fullpath), genDirectoryListKey(string(fullpath)), genDirectoryListKey(dir)},
+		store.isSuperLargeDirectory(dir), name).Err()
+
+	if err != nil {
+		return fmt.Errorf("DeleteEntry %s : %v", fullpath, err)
+	}
+
+	return nil
+}
+
+func (store *UniversalRedisLuaStore) DeleteFolderChildren(ctx context.Context, fullpath util.FullPath) (err error) {
+
+	if store.isSuperLargeDirectory(string(fullpath)) {
+		return nil
+	}
+
+	err = stored_procedure.DeleteFolderChildrenScript.Run(ctx, store.Client,
+		[]string{string(fullpath)}).Err()
+
+	if err != nil {
+		return fmt.Errorf("DeleteFolderChildren %s : %v", fullpath, err)
+	}
+
+	return nil
+}
+
+func (store *UniversalRedisLuaStore) ListDirectoryPrefixedEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, prefix string, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+	return lastFileName, filer.ErrUnsupportedListDirectoryPrefixed
+}
+
+func (store *UniversalRedisLuaStore) ListDirectoryEntries(ctx context.Context, dirPath util.FullPath, startFileName string, includeStartFile bool, limit int64, eachEntryFunc filer.ListEachEntryFunc) (lastFileName string, err error) {
+
+	dirListKey := genDirectoryListKey(string(dirPath))
+
+	min := "-"
+	if startFileName != "" {
+		if includeStartFile {
+			min = "[" + startFileName
+		} else {
+			min = "(" + startFileName
+		}
+	}
+
+	members, err := store.Client.ZRangeByLex(ctx, dirListKey, &redis.ZRangeBy{
+		Min:    min,
+		Max:    "+",
+		Offset: 0,
+		Count:  limit,
+	}).Result()
+	if err != nil {
+		return lastFileName, fmt.Errorf("list %s : %v", dirPath, err)
+	}
+
+	// fetch entry meta
+	for _, fileName := range members {
+		path := util.NewFullPath(string(dirPath), fileName)
+		entry, err := store.FindEntry(ctx, path)
+		lastFileName = fileName
+		if err != nil {
+			glog.V(0).Infof("list %s : %v", path, err)
+			if err == filer_pb.ErrNotFound {
+				continue
+			}
+		} else {
+			if entry.TtlSec > 0 {
+				if entry.Attr.Crtime.Add(time.Duration(entry.TtlSec) * time.Second).Before(time.Now()) {
+					store.DeleteEntry(ctx, path)
+					continue
+				}
+			}
+			if !eachEntryFunc(entry) {
+				break
+			}
+		}
+	}
+
+	return lastFileName, err
+}
+
+func genDirectoryListKey(dir string) (dirList string) {
+	return dir + DIR_LIST_MARKER
+}
+
+func (store *UniversalRedisLuaStore) Shutdown() {
+	store.Client.Close()
+}

weed/filer/redis_lua/universal_redis_store_kv.go

@@ -0,0 +1,42 @@
+package redis_lua
+
+import (
+	"context"
+	"fmt"
+
+	"github.com/chrislusf/seaweedfs/weed/filer"
+	"github.com/go-redis/redis/v8"
+)
+
+func (store *UniversalRedisLuaStore) KvPut(ctx context.Context, key []byte, value []byte) (err error) {
+
+	_, err = store.Client.Set(ctx, string(key), value, 0).Result()
+
+	if err != nil {
+		return fmt.Errorf("kv put: %v", err)
+	}
+
+	return nil
+}
+
+func (store *UniversalRedisLuaStore) KvGet(ctx context.Context, key []byte) (value []byte, err error) {
+
+	data, err := store.Client.Get(ctx, string(key)).Result()
+
+	if err == redis.Nil {
+		return nil, filer.ErrKvNotFound
+	}
+
+	return []byte(data), err
+}
+
+func (store *UniversalRedisLuaStore) KvDelete(ctx context.Context, key []byte) (err error) {
+
+	_, err = store.Client.Del(ctx, string(key)).Result()
+
+	if err != nil {
+		return fmt.Errorf("kv delete: %v", err)
+	}
+
+	return nil
+}

weed/filer/rocksdb/rocksdb_store_test.go

@@ -16,8 +16,7 @@ import (
 
 func TestCreateAndFind(t *testing.T) {
 	testFiler := filer.NewFiler(nil, nil, "", 0, "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_test")
-	defer os.RemoveAll(dir)
+	dir := t.TempDir()
 	store := &RocksDBStore{}
 	store.initialize(dir)
 	testFiler.SetStore(store)
@@ -35,7 +34,7 @@ func TestCreateAndFind(t *testing.T) {
 		},
 	}
 
-	if err := testFiler.CreateEntry(ctx, entry1, false, false, nil); err != nil {
+	if err := testFiler.CreateEntry(ctx, entry1, false, false, nil, false); err != nil {
 		t.Errorf("create entry %v: %v", entry1.FullPath, err)
 		return
 	}
@@ -70,8 +69,7 @@ func TestCreateAndFind(t *testing.T) {
 
 func TestEmptyRoot(t *testing.T) {
 	testFiler := filer.NewFiler(nil, nil, "", 0, "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_test2")
-	defer os.RemoveAll(dir)
+	dir := t.TempDir()
 	store := &RocksDBStore{}
 	store.initialize(dir)
 	testFiler.SetStore(store)
@@ -93,8 +91,7 @@ func TestEmptyRoot(t *testing.T) {
 
 func BenchmarkInsertEntry(b *testing.B) {
 	testFiler := filer.NewFiler(nil, nil, "", 0, "", "", "", nil)
-	dir, _ := os.MkdirTemp("", "seaweedfs_filer_bench")
-	defer os.RemoveAll(dir)
+	dir := b.TempDir()
 	store := &RocksDBStore{}
 	store.initialize(dir)
 	testFiler.SetStore(store)

weed/filer/stream.go

@@ -3,6 +3,7 @@ package filer
 import (
 	"bytes"
 	"fmt"
+	"golang.org/x/exp/slices"
 	"io"
 	"math"
 	"sort"
@@ -39,11 +40,11 @@ func isSameChunks(a, b []*filer_pb.FileChunk) bool {
 	if len(a) != len(b) {
 		return false
 	}
-	sort.Slice(a, func(i, j int) bool {
-		return strings.Compare(a[i].ETag, a[j].ETag) < 0
+	slices.SortFunc(a, func(i, j *filer_pb.FileChunk) bool {
+		return strings.Compare(i.ETag, j.ETag) < 0
 	})
-	sort.Slice(b, func(i, j int) bool {
-		return strings.Compare(b[i].ETag, b[j].ETag) < 0
+	slices.SortFunc(b, func(i, j *filer_pb.FileChunk) bool {
+		return strings.Compare(i.ETag, j.ETag) < 0
 	})
 	for i := 0; i < len(a); i++ {
 		if a[i].ETag != b[i].ETag {
@@ -62,7 +63,7 @@ func NewFileReader(filerClient filer_pb.FilerClient, entry *filer_pb.Entry) io.R
 
 func StreamContent(masterClient wdclient.HasLookupFileIdFunction, writer io.Writer, chunks []*filer_pb.FileChunk, offset int64, size int64) error {
 
-	glog.V(9).Infof("start to stream content for chunks: %+v\n", chunks)
+	glog.V(4).Infof("start to stream content for chunks: %+v", chunks)
 	chunkViews := ViewFromChunks(masterClient.GetLookupFileIdFunction(), chunks, offset, size)
 
 	fileId2Url := make(map[string][]string)
@@ -104,10 +105,12 @@ func StreamContent(masterClient wdclient.HasLookupFileIdFunction, writer io.Writ
 		}
 		stats.FilerRequestCounter.WithLabelValues("chunkDownload").Inc()
 	}
-	glog.V(4).Infof("zero [%d,%d)", offset, offset+remaining)
-	err := writeZero(writer, remaining)
-	if err != nil {
-		return fmt.Errorf("write zero [%d,%d)", offset, offset+remaining)
+	if remaining > 0 {
+		glog.V(4).Infof("zero [%d,%d)", offset, offset+remaining)
+		err := writeZero(writer, remaining)
+		if err != nil {
+			return fmt.Errorf("write zero [%d,%d)", offset, offset+remaining)
+		}
 	}
 
 	return nil
@@ -133,30 +136,30 @@ func writeZero(w io.Writer, size int64) (err error) {
 	return
 }
 
-func ReadAll(masterClient *wdclient.MasterClient, chunks []*filer_pb.FileChunk) ([]byte, error) {
-
-	buffer := bytes.Buffer{}
+func ReadAll(buffer []byte, masterClient *wdclient.MasterClient, chunks []*filer_pb.FileChunk) error {
 
 	lookupFileIdFn := func(fileId string) (targetUrls []string, err error) {
 		return masterClient.LookupFileId(fileId)
 	}
 
-	chunkViews := ViewFromChunks(lookupFileIdFn, chunks, 0, math.MaxInt64)
+	chunkViews := ViewFromChunks(lookupFileIdFn, chunks, 0, int64(len(buffer)))
+
+	idx := 0
 
 	for _, chunkView := range chunkViews {
 		urlStrings, err := lookupFileIdFn(chunkView.FileId)
 		if err != nil {
 			glog.V(1).Infof("operation LookupFileId %s failed, err: %v", chunkView.FileId, err)
-			return nil, err
+			return err
 		}
 
-		data, err := retriedFetchChunkData(urlStrings, chunkView.CipherKey, chunkView.IsGzipped, chunkView.IsFullChunk(), chunkView.Offset, int(chunkView.Size))
+		n, err := retriedFetchChunkData(buffer[idx:idx+int(chunkView.Size)], urlStrings, chunkView.CipherKey, chunkView.IsGzipped, chunkView.IsFullChunk(), chunkView.Offset)
 		if err != nil {
-			return nil, err
+			return err
 		}
-		buffer.Write(data)
+		idx += n
 	}
-	return buffer.Bytes(), nil
+	return nil
 }
 
 // ----------------  ChunkStreamReader ----------------------------------
@@ -177,8 +180,8 @@ var _ = io.ReaderAt(&ChunkStreamReader{})
 func doNewChunkStreamReader(lookupFileIdFn wdclient.LookupFileIdFunctionType, chunks []*filer_pb.FileChunk) *ChunkStreamReader {
 
 	chunkViews := ViewFromChunks(lookupFileIdFn, chunks, 0, math.MaxInt64)
-	sort.Slice(chunkViews, func(i, j int) bool {
-		return chunkViews[i].LogicOffset < chunkViews[j].LogicOffset
+	slices.SortFunc(chunkViews, func(a, b *ChunkView) bool {
+		return a.LogicOffset < b.LogicOffset
 	})
 
 	var totalSize int64

weed/filesys/dir.go

@@ -1,655 +0,0 @@
-package filesys
-
-import (
-	"bytes"
-	"context"
-	"math"
-	"os"
-	"strings"
-	"syscall"
-	"time"
-
-	"github.com/seaweedfs/fuse"
-	"github.com/seaweedfs/fuse/fs"
-
-	"github.com/chrislusf/seaweedfs/weed/filer"
-	"github.com/chrislusf/seaweedfs/weed/filesys/meta_cache"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"github.com/chrislusf/seaweedfs/weed/util"
-)
-
-type Dir struct {
-	name   string
-	wfs    *WFS
-	entry  *filer_pb.Entry
-	parent *Dir
-	id     uint64
-}
-
-var _ = fs.Node(&Dir{})
-
-var _ = fs.NodeIdentifier(&Dir{})
-var _ = fs.NodeCreater(&Dir{})
-var _ = fs.NodeMknoder(&Dir{})
-var _ = fs.NodeMkdirer(&Dir{})
-var _ = fs.NodeFsyncer(&Dir{})
-var _ = fs.NodeRequestLookuper(&Dir{})
-var _ = fs.HandleReadDirAller(&Dir{})
-var _ = fs.NodeRemover(&Dir{})
-var _ = fs.NodeRenamer(&Dir{})
-var _ = fs.NodeSetattrer(&Dir{})
-var _ = fs.NodeGetxattrer(&Dir{})
-var _ = fs.NodeSetxattrer(&Dir{})
-var _ = fs.NodeRemovexattrer(&Dir{})
-var _ = fs.NodeListxattrer(&Dir{})
-var _ = fs.NodeForgetter(&Dir{})
-
-func (dir *Dir) Id() uint64 {
-	if dir.parent == nil {
-		return 1
-	}
-	return dir.id
-}
-
-func (dir *Dir) Attr(ctx context.Context, attr *fuse.Attr) error {
-
-	entry, err := dir.maybeLoadEntry()
-	if err != nil {
-		glog.V(3).Infof("dir Attr %s, err: %+v", dir.FullPath(), err)
-		return err
-	}
-
-	// https://github.com/bazil/fuse/issues/196
-	attr.Valid = time.Second
-	attr.Inode = dir.Id()
-	attr.Mode = os.FileMode(entry.Attributes.FileMode) | os.ModeDir
-	attr.Mtime = time.Unix(entry.Attributes.Mtime, 0)
-	attr.Crtime = time.Unix(entry.Attributes.Crtime, 0)
-	attr.Ctime = time.Unix(entry.Attributes.Mtime, 0)
-	attr.Atime = time.Unix(entry.Attributes.Mtime, 0)
-	attr.Gid = entry.Attributes.Gid
-	attr.Uid = entry.Attributes.Uid
-
-	if dir.FullPath() == dir.wfs.option.FilerMountRootPath {
-		attr.BlockSize = blockSize
-	}
-
-	glog.V(4).Infof("dir Attr %s, attr: %+v", dir.FullPath(), attr)
-
-	return nil
-}
-
-func (dir *Dir) Getxattr(ctx context.Context, req *fuse.GetxattrRequest, resp *fuse.GetxattrResponse) error {
-
-	glog.V(4).Infof("dir Getxattr %s", dir.FullPath())
-
-	entry, err := dir.maybeLoadEntry()
-	if err != nil {
-		return err
-	}
-
-	return getxattr(entry, req, resp)
-}
-
-func (dir *Dir) Fsync(ctx context.Context, req *fuse.FsyncRequest) error {
-	// fsync works at OS level
-	// write the file chunks to the filerGrpcAddress
-	glog.V(3).Infof("dir %s fsync %+v", dir.FullPath(), req)
-
-	return nil
-}
-
-func (dir *Dir) newFile(name string, fileMode os.FileMode) fs.Node {
-
-	fileFullPath := util.NewFullPath(dir.FullPath(), name)
-	fileId := fileFullPath.AsInode(fileMode)
-	dir.wfs.handlesLock.Lock()
-	existingHandle, found := dir.wfs.handles[fileId]
-	dir.wfs.handlesLock.Unlock()
-
-	if found {
-		glog.V(4).Infof("newFile found opened file handle: %+v", fileFullPath)
-		return existingHandle.f
-	}
-	return &File{
-		Name: name,
-		dir:  dir,
-		wfs:  dir.wfs,
-		id:   fileId,
-	}
-}
-
-func (dir *Dir) newDirectory(fullpath util.FullPath) fs.Node {
-
-	return &Dir{name: fullpath.Name(), wfs: dir.wfs, parent: dir, id: fullpath.AsInode(os.ModeDir)}
-
-}
-
-func (dir *Dir) Create(ctx context.Context, req *fuse.CreateRequest,
-	resp *fuse.CreateResponse) (fs.Node, fs.Handle, error) {
-
-	if err := checkName(req.Name); err != nil {
-		return nil, nil, err
-	}
-
-	exclusive := req.Flags&fuse.OpenExclusive != 0
-	isDirectory := req.Mode&os.ModeDir > 0
-
-	if exclusive || isDirectory {
-		_, err := dir.doCreateEntry(req.Name, req.Mode, req.Uid, req.Gid, exclusive)
-		if err != nil {
-			return nil, nil, err
-		}
-	}
-	var node fs.Node
-	if isDirectory {
-		node = dir.newDirectory(util.NewFullPath(dir.FullPath(), req.Name))
-		return node, node, nil
-	}
-
-	node = dir.newFile(req.Name, req.Mode)
-	file := node.(*File)
-	file.entry = &filer_pb.Entry{
-		Name:        req.Name,
-		IsDirectory: req.Mode&os.ModeDir > 0,
-		Attributes: &filer_pb.FuseAttributes{
-			Mtime:       time.Now().Unix(),
-			Crtime:      time.Now().Unix(),
-			FileMode:    uint32(req.Mode &^ dir.wfs.option.Umask),
-			Uid:         req.Uid,
-			Gid:         req.Gid,
-			Collection:  dir.wfs.option.Collection,
-			Replication: dir.wfs.option.Replication,
-			TtlSec:      dir.wfs.option.TtlSec,
-		},
-	}
-	file.dirtyMetadata = true
-	fh := dir.wfs.AcquireHandle(file, req.Uid, req.Gid)
-	return file, fh, nil
-
-}
-
-func (dir *Dir) Mknod(ctx context.Context, req *fuse.MknodRequest) (fs.Node, error) {
-
-	if err := checkName(req.Name); err != nil {
-		return nil, err
-	}
-
-	glog.V(3).Infof("dir %s Mknod %+v", dir.FullPath(), req)
-
-	_, err := dir.doCreateEntry(req.Name, req.Mode, req.Uid, req.Gid, false)
-
-	if err != nil {
-		return nil, err
-	}
-	var node fs.Node
-	node = dir.newFile(req.Name, req.Mode)
-	return node, nil
-}
-
-func (dir *Dir) doCreateEntry(name string, mode os.FileMode, uid, gid uint32, exclusive bool) (*filer_pb.CreateEntryRequest, error) {
-	dirFullPath := dir.FullPath()
-	request := &filer_pb.CreateEntryRequest{
-		Directory: dirFullPath,
-		Entry: &filer_pb.Entry{
-			Name:        name,
-			IsDirectory: mode&os.ModeDir > 0,
-			Attributes: &filer_pb.FuseAttributes{
-				Mtime:       time.Now().Unix(),
-				Crtime:      time.Now().Unix(),
-				FileMode:    uint32(mode &^ dir.wfs.option.Umask),
-				Uid:         uid,
-				Gid:         gid,
-				Collection:  dir.wfs.option.Collection,
-				Replication: dir.wfs.option.Replication,
-				TtlSec:      dir.wfs.option.TtlSec,
-			},
-		},
-		OExcl:      exclusive,
-		Signatures: []int32{dir.wfs.signature},
-	}
-	glog.V(1).Infof("create %s/%s", dirFullPath, name)
-
-	err := dir.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
-
-		dir.wfs.mapPbIdFromLocalToFiler(request.Entry)
-		defer dir.wfs.mapPbIdFromFilerToLocal(request.Entry)
-
-		if err := filer_pb.CreateEntry(client, request); err != nil {
-			if strings.Contains(err.Error(), "EEXIST") {
-				return fuse.EEXIST
-			}
-			glog.V(0).Infof("create %s/%s: %v", dirFullPath, name, err)
-			return fuse.EIO
-		}
-
-		if err := dir.wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry)); err != nil {
-			glog.Errorf("local InsertEntry dir %s/%s: %v", dirFullPath, name, err)
-			return fuse.EIO
-		}
-
-		return nil
-	})
-	return request, err
-}
-
-func (dir *Dir) Mkdir(ctx context.Context, req *fuse.MkdirRequest) (fs.Node, error) {
-
-	if err := checkName(req.Name); err != nil {
-		return nil, err
-	}
-
-	glog.V(4).Infof("mkdir %s: %s", dir.FullPath(), req.Name)
-
-	newEntry := &filer_pb.Entry{
-		Name:        req.Name,
-		IsDirectory: true,
-		Attributes: &filer_pb.FuseAttributes{
-			Mtime:    time.Now().Unix(),
-			Crtime:   time.Now().Unix(),
-			FileMode: uint32(req.Mode &^ dir.wfs.option.Umask),
-			Uid:      req.Uid,
-			Gid:      req.Gid,
-		},
-	}
-
-	dirFullPath := dir.FullPath()
-
-	err := dir.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
-
-		dir.wfs.mapPbIdFromLocalToFiler(newEntry)
-		defer dir.wfs.mapPbIdFromFilerToLocal(newEntry)
-
-		request := &filer_pb.CreateEntryRequest{
-			Directory:  dirFullPath,
-			Entry:      newEntry,
-			Signatures: []int32{dir.wfs.signature},
-		}
-
-		glog.V(1).Infof("mkdir: %v", request)
-		if err := filer_pb.CreateEntry(client, request); err != nil {
-			glog.V(0).Infof("mkdir %s/%s: %v", dirFullPath, req.Name, err)
-			return err
-		}
-
-		if err := dir.wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry)); err != nil {
-			glog.Errorf("local mkdir dir %s/%s: %v", dirFullPath, req.Name, err)
-			return fuse.EIO
-		}
-
-		return nil
-	})
-
-	if err == nil {
-		node := dir.newDirectory(util.NewFullPath(dirFullPath, req.Name))
-
-		return node, nil
-	}
-
-	glog.V(0).Infof("mkdir %s/%s: %v", dirFullPath, req.Name, err)
-
-	return nil, fuse.EIO
-}
-
-func (dir *Dir) Lookup(ctx context.Context, req *fuse.LookupRequest, resp *fuse.LookupResponse) (node fs.Node, err error) {
-
-	if err := checkName(req.Name); err != nil {
-		return nil, err
-	}
-
-	dirPath := util.FullPath(dir.FullPath())
-	// glog.V(4).Infof("dir Lookup %s: %s by %s", dirPath, req.Name, req.Header.String())
-
-	fullFilePath := dirPath.Child(req.Name)
-	visitErr := meta_cache.EnsureVisited(dir.wfs.metaCache, dir.wfs, dirPath)
-	if visitErr != nil {
-		glog.Errorf("dir Lookup %s: %v", dirPath, visitErr)
-		return nil, fuse.EIO
-	}
-	localEntry, cacheErr := dir.wfs.metaCache.FindEntry(context.Background(), fullFilePath)
-	if cacheErr == filer_pb.ErrNotFound {
-		return nil, fuse.ENOENT
-	}
-
-	if localEntry == nil {
-		// glog.V(3).Infof("dir Lookup cache miss %s", fullFilePath)
-		entry, err := filer_pb.GetEntry(dir.wfs, fullFilePath)
-		if err != nil {
-			glog.V(1).Infof("dir GetEntry %s: %v", fullFilePath, err)
-			return nil, fuse.ENOENT
-		}
-		localEntry = filer.FromPbEntry(string(dirPath), entry)
-	} else {
-		glog.V(4).Infof("dir Lookup cache hit %s", fullFilePath)
-	}
-
-	if localEntry != nil {
-		if localEntry.IsDirectory() {
-			node = dir.newDirectory(fullFilePath)
-		} else {
-			node = dir.newFile(req.Name, localEntry.Attr.Mode)
-		}
-
-		// resp.EntryValid = time.Second
-		resp.Attr.Valid = time.Second
-		resp.Attr.Size = localEntry.FileSize
-		resp.Attr.Mtime = localEntry.Attr.Mtime
-		resp.Attr.Crtime = localEntry.Attr.Crtime
-		resp.Attr.Mode = localEntry.Attr.Mode
-		resp.Attr.Gid = localEntry.Attr.Gid
-		resp.Attr.Uid = localEntry.Attr.Uid
-		if localEntry.HardLinkCounter > 0 {
-			resp.Attr.Nlink = uint32(localEntry.HardLinkCounter)
-		}
-
-		return node, nil
-	}
-
-	glog.V(4).Infof("not found dir GetEntry %s: %v", fullFilePath, err)
-	return nil, fuse.ENOENT
-}
-
-func (dir *Dir) ReadDirAll(ctx context.Context) (ret []fuse.Dirent, err error) {
-
-	dirPath := util.FullPath(dir.FullPath())
-	glog.V(4).Infof("dir ReadDirAll %s", dirPath)
-
-	processEachEntryFn := func(entry *filer.Entry, isLast bool) {
-		if entry.IsDirectory() {
-			dirent := fuse.Dirent{Name: entry.Name(), Type: fuse.DT_Dir, Inode: dirPath.Child(entry.Name()).AsInode(os.ModeDir)}
-			ret = append(ret, dirent)
-		} else {
-			dirent := fuse.Dirent{Name: entry.Name(), Type: findFileType(uint16(entry.Attr.Mode)), Inode: dirPath.Child(entry.Name()).AsInode(entry.Attr.Mode)}
-			ret = append(ret, dirent)
-		}
-	}
-
-	if err = meta_cache.EnsureVisited(dir.wfs.metaCache, dir.wfs, dirPath); err != nil {
-		glog.Errorf("dir ReadDirAll %s: %v", dirPath, err)
-		return nil, fuse.EIO
-	}
-	listErr := dir.wfs.metaCache.ListDirectoryEntries(context.Background(), dirPath, "", false, int64(math.MaxInt32), func(entry *filer.Entry) bool {
-		processEachEntryFn(entry, false)
-		return true
-	})
-	if listErr != nil {
-		glog.Errorf("list meta cache: %v", listErr)
-		return nil, fuse.EIO
-	}
-
-	// create proper . and .. directories
-	ret = append(ret, fuse.Dirent{
-		Inode: dirPath.AsInode(os.ModeDir),
-		Name:  ".",
-		Type:  fuse.DT_Dir,
-	})
-
-	// return the correct parent inode for the mount root
-	var inode uint64
-	if string(dirPath) == dir.wfs.option.FilerMountRootPath {
-		inode = dir.wfs.option.MountParentInode
-	} else {
-		inode = util.FullPath(dir.parent.FullPath()).AsInode(os.ModeDir)
-	}
-
-	ret = append(ret, fuse.Dirent{
-		Inode: inode,
-		Name:  "..",
-		Type:  fuse.DT_Dir,
-	})
-
-	return
-}
-
-func findFileType(mode uint16) fuse.DirentType {
-	switch mode & (syscall.S_IFMT & 0xffff) {
-	case syscall.S_IFSOCK:
-		return fuse.DT_Socket
-	case syscall.S_IFLNK:
-		return fuse.DT_Link
-	case syscall.S_IFREG:
-		return fuse.DT_File
-	case syscall.S_IFBLK:
-		return fuse.DT_Block
-	case syscall.S_IFDIR:
-		return fuse.DT_Dir
-	case syscall.S_IFCHR:
-		return fuse.DT_Char
-	case syscall.S_IFIFO:
-		return fuse.DT_FIFO
-	}
-	return fuse.DT_File
-}
-
-func (dir *Dir) Remove(ctx context.Context, req *fuse.RemoveRequest) error {
-
-	parentHasPermission := false
-	parentEntry, err := dir.maybeLoadEntry()
-	if err != nil {
-		return err
-	}
-	if err := checkPermission(parentEntry, req.Uid, req.Gid, true); err == nil {
-		parentHasPermission = true
-	}
-
-	dirFullPath := dir.FullPath()
-	filePath := util.NewFullPath(dirFullPath, req.Name)
-	entry, err := filer_pb.GetEntry(dir.wfs, filePath)
-	if err != nil {
-		return err
-	}
-	if !parentHasPermission {
-		if err := checkPermission(entry, req.Uid, req.Gid, true); err != nil {
-			return err
-		}
-	}
-
-	if !req.Dir {
-		return dir.removeOneFile(entry, req)
-	}
-
-	return dir.removeFolder(entry, req)
-
-}
-
-func (dir *Dir) removeOneFile(entry *filer_pb.Entry, req *fuse.RemoveRequest) error {
-
-	dirFullPath := dir.FullPath()
-	filePath := util.NewFullPath(dirFullPath, req.Name)
-
-	// first, ensure the filer store can correctly delete
-	glog.V(3).Infof("remove file: %v", req)
-	isDeleteData := entry != nil && entry.HardLinkCounter <= 1
-	err := filer_pb.Remove(dir.wfs, dirFullPath, req.Name, isDeleteData, false, false, false, []int32{dir.wfs.signature})
-	if err != nil {
-		glog.V(3).Infof("not found remove file %s: %v", filePath, err)
-		return fuse.ENOENT
-	}
-
-	// then, delete meta cache and fsNode cache
-	if err = dir.wfs.metaCache.DeleteEntry(context.Background(), filePath); err != nil {
-		glog.V(3).Infof("local DeleteEntry %s: %v", filePath, err)
-		return fuse.ESTALE
-	}
-
-	// remove current file handle if any
-	dir.wfs.handlesLock.Lock()
-	defer dir.wfs.handlesLock.Unlock()
-	inodeId := filePath.AsInode(0)
-	if fh, ok := dir.wfs.handles[inodeId]; ok {
-		delete(dir.wfs.handles, inodeId)
-		fh.isDeleted = true
-	}
-
-	return nil
-
-}
-
-func (dir *Dir) removeFolder(entry *filer_pb.Entry, req *fuse.RemoveRequest) error {
-
-	dirFullPath := dir.FullPath()
-
-	glog.V(3).Infof("remove directory entry: %v", req)
-	ignoreRecursiveErr := true // ignore recursion error since the OS should manage it
-	err := filer_pb.Remove(dir.wfs, dirFullPath, req.Name, true, true, ignoreRecursiveErr, false, []int32{dir.wfs.signature})
-	if err != nil {
-		glog.V(0).Infof("remove %s/%s: %v", dirFullPath, req.Name, err)
-		if strings.Contains(err.Error(), filer.MsgFailDelNonEmptyFolder) {
-			return fuse.EEXIST
-		}
-		return fuse.ENOENT
-	}
-
-	t := util.NewFullPath(dirFullPath, req.Name)
-	dir.wfs.metaCache.DeleteEntry(context.Background(), t)
-
-	return nil
-
-}
-
-func (dir *Dir) Setattr(ctx context.Context, req *fuse.SetattrRequest, resp *fuse.SetattrResponse) error {
-
-	glog.V(4).Infof("%v dir setattr %+v mode=%d", dir.FullPath(), req, req.Mode)
-
-	entry, err := dir.maybeLoadEntry()
-	if err != nil {
-		return err
-	}
-
-	if req.Valid.Mode() {
-		entry.Attributes.FileMode = uint32(req.Mode)
-	}
-
-	if req.Valid.Uid() {
-		entry.Attributes.Uid = req.Uid
-	}
-
-	if req.Valid.Gid() {
-		entry.Attributes.Gid = req.Gid
-	}
-
-	if req.Valid.Mtime() {
-		entry.Attributes.Mtime = req.Mtime.Unix()
-	}
-
-	entry.Attributes.Mtime = time.Now().Unix()
-
-	return dir.saveEntry(entry)
-
-}
-
-func (dir *Dir) Setxattr(ctx context.Context, req *fuse.SetxattrRequest) error {
-
-	glog.V(4).Infof("dir Setxattr %s: %s", dir.FullPath(), req.Name)
-
-	entry, err := dir.maybeLoadEntry()
-	if err != nil {
-		return err
-	}
-
-	if err := setxattr(entry, req); err != nil {
-		return err
-	}
-
-	return dir.saveEntry(entry)
-
-}
-
-func (dir *Dir) Removexattr(ctx context.Context, req *fuse.RemovexattrRequest) error {
-
-	glog.V(4).Infof("dir Removexattr %s: %s", dir.FullPath(), req.Name)
-
-	entry, err := dir.maybeLoadEntry()
-	if err != nil {
-		return err
-	}
-
-	if err := removexattr(entry, req); err != nil {
-		return err
-	}
-
-	return dir.saveEntry(entry)
-
-}
-
-func (dir *Dir) Listxattr(ctx context.Context, req *fuse.ListxattrRequest, resp *fuse.ListxattrResponse) error {
-
-	glog.V(4).Infof("dir Listxattr %s", dir.FullPath())
-
-	entry, err := dir.maybeLoadEntry()
-	if err != nil {
-		return err
-	}
-
-	if err := listxattr(entry, req, resp); err != nil {
-		return err
-	}
-
-	return nil
-
-}
-
-func (dir *Dir) Forget() {
-	glog.V(4).Infof("Forget dir %s", dir.FullPath())
-}
-
-func (dir *Dir) maybeLoadEntry() (*filer_pb.Entry, error) {
-	parentDirPath, name := util.FullPath(dir.FullPath()).DirAndName()
-	return dir.wfs.maybeLoadEntry(parentDirPath, name)
-}
-
-func (dir *Dir) saveEntry(entry *filer_pb.Entry) error {
-
-	parentDir, name := util.FullPath(dir.FullPath()).DirAndName()
-
-	return dir.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
-
-		dir.wfs.mapPbIdFromLocalToFiler(entry)
-		defer dir.wfs.mapPbIdFromFilerToLocal(entry)
-
-		request := &filer_pb.UpdateEntryRequest{
-			Directory:  parentDir,
-			Entry:      entry,
-			Signatures: []int32{dir.wfs.signature},
-		}
-
-		glog.V(1).Infof("save dir entry: %v", request)
-		_, err := client.UpdateEntry(context.Background(), request)
-		if err != nil {
-			glog.Errorf("UpdateEntry dir %s/%s: %v", parentDir, name, err)
-			return fuse.EIO
-		}
-
-		if err := dir.wfs.metaCache.UpdateEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry)); err != nil {
-			glog.Errorf("UpdateEntry dir %s/%s: %v", parentDir, name, err)
-			return fuse.ESTALE
-		}
-
-		return nil
-	})
-}
-
-func (dir *Dir) FullPath() string {
-	var parts []string
-	for p := dir; p != nil; p = p.parent {
-		if strings.HasPrefix(p.name, "/") {
-			if len(p.name) > 1 {
-				parts = append(parts, p.name[1:])
-			}
-		} else {
-			parts = append(parts, p.name)
-		}
-	}
-
-	if len(parts) == 0 {
-		return "/"
-	}
-
-	var buf bytes.Buffer
-	for i := len(parts) - 1; i >= 0; i-- {
-		buf.WriteString("/")
-		buf.WriteString(parts[i])
-	}
-	return buf.String()
-}

weed/filesys/dir_link.go

@@ -1,169 +0,0 @@
-package filesys
-
-import (
-	"context"
-	"github.com/chrislusf/seaweedfs/weed/util"
-	"os"
-	"syscall"
-	"time"
-
-	"github.com/chrislusf/seaweedfs/weed/filer"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"github.com/seaweedfs/fuse"
-	"github.com/seaweedfs/fuse/fs"
-)
-
-var _ = fs.NodeLinker(&Dir{})
-var _ = fs.NodeSymlinker(&Dir{})
-var _ = fs.NodeReadlinker(&File{})
-
-const (
-	HARD_LINK_MARKER = '\x01'
-)
-
-func (dir *Dir) Link(ctx context.Context, req *fuse.LinkRequest, old fs.Node) (fs.Node, error) {
-
-	if err := checkName(req.NewName); err != nil {
-		return nil, err
-	}
-
-	oldFile, ok := old.(*File)
-	if !ok {
-		glog.Errorf("old node is not a file: %+v", old)
-	}
-
-	glog.V(4).Infof("Link: %v/%v -> %v/%v", oldFile.dir.FullPath(), oldFile.Name, dir.FullPath(), req.NewName)
-
-	oldEntry, err := oldFile.maybeLoadEntry(ctx)
-	if err != nil {
-		return nil, err
-	}
-
-	if oldEntry == nil {
-		return nil, fuse.EIO
-	}
-
-	// update old file to hardlink mode
-	if len(oldEntry.HardLinkId) == 0 {
-		oldEntry.HardLinkId = append(util.RandomBytes(16), HARD_LINK_MARKER)
-		oldEntry.HardLinkCounter = 1
-	}
-	oldEntry.HardLinkCounter++
-	updateOldEntryRequest := &filer_pb.UpdateEntryRequest{
-		Directory:  oldFile.dir.FullPath(),
-		Entry:      oldEntry,
-		Signatures: []int32{dir.wfs.signature},
-	}
-
-	// CreateLink 1.2 : update new file to hardlink mode
-	oldEntry.Attributes.Mtime = time.Now().Unix()
-	request := &filer_pb.CreateEntryRequest{
-		Directory: dir.FullPath(),
-		Entry: &filer_pb.Entry{
-			Name:            req.NewName,
-			IsDirectory:     false,
-			Attributes:      oldEntry.Attributes,
-			Chunks:          oldEntry.Chunks,
-			Extended:        oldEntry.Extended,
-			HardLinkId:      oldEntry.HardLinkId,
-			HardLinkCounter: oldEntry.HardLinkCounter,
-		},
-		Signatures: []int32{dir.wfs.signature},
-	}
-
-	// apply changes to the filer, and also apply to local metaCache
-	err = dir.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
-
-		dir.wfs.mapPbIdFromLocalToFiler(request.Entry)
-		defer dir.wfs.mapPbIdFromFilerToLocal(request.Entry)
-
-		if err := filer_pb.UpdateEntry(client, updateOldEntryRequest); err != nil {
-			glog.V(0).Infof("Link %v/%v -> %s/%s: %v", oldFile.dir.FullPath(), oldFile.Name, dir.FullPath(), req.NewName, err)
-			return fuse.EIO
-		}
-		dir.wfs.metaCache.UpdateEntry(context.Background(), filer.FromPbEntry(updateOldEntryRequest.Directory, updateOldEntryRequest.Entry))
-
-		if err := filer_pb.CreateEntry(client, request); err != nil {
-			glog.V(0).Infof("Link %v/%v -> %s/%s: %v", oldFile.dir.FullPath(), oldFile.Name, dir.FullPath(), req.NewName, err)
-			return fuse.EIO
-		}
-		dir.wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry))
-
-		return nil
-	})
-
-	if err != nil {
-		return nil, fuse.EIO
-	}
-
-	// create new file node
-	newNode := dir.newFile(req.NewName, 0)
-	newFile := newNode.(*File)
-
-	return newFile, err
-
-}
-
-func (dir *Dir) Symlink(ctx context.Context, req *fuse.SymlinkRequest) (fs.Node, error) {
-
-	if err := checkName(req.NewName); err != nil {
-		return nil, err
-	}
-
-	glog.V(4).Infof("Symlink: %v/%v to %v", dir.FullPath(), req.NewName, req.Target)
-
-	request := &filer_pb.CreateEntryRequest{
-		Directory: dir.FullPath(),
-		Entry: &filer_pb.Entry{
-			Name:        req.NewName,
-			IsDirectory: false,
-			Attributes: &filer_pb.FuseAttributes{
-				Mtime:         time.Now().Unix(),
-				Crtime:        time.Now().Unix(),
-				FileMode:      uint32((os.FileMode(0777) | os.ModeSymlink) &^ dir.wfs.option.Umask),
-				Uid:           req.Uid,
-				Gid:           req.Gid,
-				SymlinkTarget: req.Target,
-			},
-		},
-		Signatures: []int32{dir.wfs.signature},
-	}
-
-	err := dir.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
-
-		dir.wfs.mapPbIdFromLocalToFiler(request.Entry)
-		defer dir.wfs.mapPbIdFromFilerToLocal(request.Entry)
-
-		if err := filer_pb.CreateEntry(client, request); err != nil {
-			glog.V(0).Infof("symlink %s/%s: %v", dir.FullPath(), req.NewName, err)
-			return fuse.EIO
-		}
-
-		dir.wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry))
-
-		return nil
-	})
-
-	symlink := dir.newFile(req.NewName, os.ModeSymlink)
-
-	return symlink, err
-
-}
-
-func (file *File) Readlink(ctx context.Context, req *fuse.ReadlinkRequest) (string, error) {
-
-	entry, err := file.maybeLoadEntry(ctx)
-	if err != nil {
-		return "", err
-	}
-
-	if os.FileMode(entry.Attributes.FileMode)&os.ModeSymlink == 0 {
-		return "", fuse.Errno(syscall.EINVAL)
-	}
-
-	glog.V(4).Infof("Readlink: %v/%v => %v", file.dir.FullPath(), file.Name, entry.Attributes.SymlinkTarget)
-
-	return entry.Attributes.SymlinkTarget, nil
-
-}

weed/filesys/dir_rename.go

@@ -1,149 +0,0 @@
-package filesys
-
-import (
-	"context"
-	"github.com/chrislusf/seaweedfs/weed/filer"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"github.com/chrislusf/seaweedfs/weed/util"
-	"github.com/seaweedfs/fuse"
-	"github.com/seaweedfs/fuse/fs"
-	"io"
-	"os"
-	"strings"
-)
-
-func (dir *Dir) Rename(ctx context.Context, req *fuse.RenameRequest, newDirectory fs.Node) error {
-
-	if err := checkName(req.NewName); err != nil {
-		return err
-	}
-	if err := checkName(req.OldName); err != nil {
-		return err
-	}
-
-	newDir := newDirectory.(*Dir)
-
-	newPath := util.NewFullPath(newDir.FullPath(), req.NewName)
-	oldPath := util.NewFullPath(dir.FullPath(), req.OldName)
-
-	glog.V(4).Infof("dir Rename %s => %s", oldPath, newPath)
-
-	// update remote filer
-	err := dir.wfs.WithFilerClient(true, func(client filer_pb.SeaweedFilerClient) error {
-		ctx, cancel := context.WithCancel(context.Background())
-		defer cancel()
-
-		request := &filer_pb.StreamRenameEntryRequest{
-			OldDirectory: dir.FullPath(),
-			OldName:      req.OldName,
-			NewDirectory: newDir.FullPath(),
-			NewName:      req.NewName,
-			Signatures:   []int32{dir.wfs.signature},
-		}
-
-		stream, err := client.StreamRenameEntry(ctx, request)
-		if err != nil {
-			glog.Errorf("dir AtomicRenameEntry %s => %s : %v", oldPath, newPath, err)
-			return fuse.EIO
-		}
-
-		for {
-			resp, recvErr := stream.Recv()
-			if recvErr != nil {
-				if recvErr == io.EOF {
-					break
-				} else {
-					glog.V(0).Infof("dir Rename %s => %s receive: %v", oldPath, newPath, recvErr)
-					if strings.Contains(recvErr.Error(), "not empty") {
-						return fuse.EEXIST
-					}
-					if strings.Contains(recvErr.Error(), "not directory") {
-						return fuse.ENOTDIR
-					}
-					return recvErr
-				}
-			}
-
-			if err = dir.handleRenameResponse(ctx, resp); err != nil {
-				glog.V(0).Infof("dir Rename %s => %s : %v", oldPath, newPath, err)
-				return err
-			}
-
-		}
-
-		return nil
-
-	})
-
-	return err
-}
-
-func (dir *Dir) handleRenameResponse(ctx context.Context, resp *filer_pb.StreamRenameEntryResponse) error {
-	// comes from filer StreamRenameEntry, can only be create or delete entry
-
-	if resp.EventNotification.NewEntry != nil {
-		// with new entry, the old entry name also exists. This is the first step to create new entry
-		newEntry := filer.FromPbEntry(resp.EventNotification.NewParentPath, resp.EventNotification.NewEntry)
-		if err := dir.wfs.metaCache.AtomicUpdateEntryFromFiler(ctx, "", newEntry); err != nil {
-			return err
-		}
-
-		oldParent, newParent := util.FullPath(resp.Directory), util.FullPath(resp.EventNotification.NewParentPath)
-		oldName, newName := resp.EventNotification.OldEntry.Name, resp.EventNotification.NewEntry.Name
-
-		entryFileMode := newEntry.Attr.Mode
-
-		oldPath := oldParent.Child(oldName)
-		newPath := newParent.Child(newName)
-		oldFsNode := NodeWithId(oldPath.AsInode(entryFileMode))
-		newFsNode := NodeWithId(newPath.AsInode(entryFileMode))
-		newDirNode, found := dir.wfs.Server.FindInternalNode(NodeWithId(newParent.AsInode(os.ModeDir)))
-		var newDir *Dir
-		if found {
-			newDir = newDirNode.(*Dir)
-		}
-		dir.wfs.Server.InvalidateInternalNode(oldFsNode, newFsNode, func(internalNode fs.Node) {
-			if file, ok := internalNode.(*File); ok {
-				glog.V(4).Infof("internal file node %s", oldParent.Child(oldName))
-				file.Name = newName
-				file.id = uint64(newFsNode)
-				if found {
-					file.dir = newDir
-				}
-			}
-			if dir, ok := internalNode.(*Dir); ok {
-				glog.V(4).Infof("internal dir node %s", oldParent.Child(oldName))
-				dir.name = newName
-				dir.id = uint64(newFsNode)
-				if found {
-					dir.parent = newDir
-				}
-			}
-		})
-
-		// change file handle
-		if !newEntry.IsDirectory() {
-			inodeId := oldPath.AsInode(entryFileMode)
-			dir.wfs.handlesLock.Lock()
-			if existingHandle, found := dir.wfs.handles[inodeId]; found && existingHandle != nil {
-				glog.V(4).Infof("opened file handle %s => %s", oldPath, newPath)
-				delete(dir.wfs.handles, inodeId)
-				existingHandle.handle = newPath.AsInode(entryFileMode)
-				existingHandle.f.entry.Name = newName
-				existingHandle.f.id = newPath.AsInode(entryFileMode)
-				dir.wfs.handles[newPath.AsInode(entryFileMode)] = existingHandle
-			}
-			dir.wfs.handlesLock.Unlock()
-		}
-
-	} else if resp.EventNotification.OldEntry != nil {
-		// without new entry, only old entry name exists. This is the second step to delete old entry
-		if err := dir.wfs.metaCache.AtomicUpdateEntryFromFiler(ctx, util.NewFullPath(resp.Directory, resp.EventNotification.OldEntry.Name), nil); err != nil {
-			return err
-		}
-	}
-
-	return nil
-
-}

weed/filesys/dirty_pages_chunked.go

@@ -1,100 +0,0 @@
-package filesys
-
-import (
-	"fmt"
-	"github.com/chrislusf/seaweedfs/weed/filesys/page_writer"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"io"
-	"sync"
-	"time"
-)
-
-type ChunkedDirtyPages struct {
-	fh             *FileHandle
-	writeWaitGroup sync.WaitGroup
-	chunkAddLock   sync.Mutex
-	lastErr        error
-	collection     string
-	replication    string
-	uploadPipeline *page_writer.UploadPipeline
-	hasWrites      bool
-}
-
-var (
-	_ = page_writer.DirtyPages(&ChunkedDirtyPages{})
-)
-
-func newMemoryChunkPages(fh *FileHandle, chunkSize int64) *ChunkedDirtyPages {
-
-	dirtyPages := &ChunkedDirtyPages{
-		fh: fh,
-	}
-
-	dirtyPages.uploadPipeline = page_writer.NewUploadPipeline(fh.f.wfs.concurrentWriters, chunkSize, dirtyPages.saveChunkedFileIntevalToStorage, fh.f.wfs.option.ConcurrentWriters)
-
-	return dirtyPages
-}
-
-func (pages *ChunkedDirtyPages) AddPage(offset int64, data []byte) {
-	pages.hasWrites = true
-
-	glog.V(4).Infof("%v memory AddPage [%d, %d)", pages.fh.f.fullpath(), offset, offset+int64(len(data)))
-	pages.uploadPipeline.SaveDataAt(data, offset)
-
-	return
-}
-
-func (pages *ChunkedDirtyPages) FlushData() error {
-	if !pages.hasWrites {
-		return nil
-	}
-	pages.uploadPipeline.FlushAll()
-	if pages.lastErr != nil {
-		return fmt.Errorf("flush data: %v", pages.lastErr)
-	}
-	return nil
-}
-
-func (pages *ChunkedDirtyPages) ReadDirtyDataAt(data []byte, startOffset int64) (maxStop int64) {
-	if !pages.hasWrites {
-		return
-	}
-	return pages.uploadPipeline.MaybeReadDataAt(data, startOffset)
-}
-
-func (pages *ChunkedDirtyPages) GetStorageOptions() (collection, replication string) {
-	return pages.collection, pages.replication
-}
-
-func (pages *ChunkedDirtyPages) saveChunkedFileIntevalToStorage(reader io.Reader, offset int64, size int64, cleanupFn func()) {
-
-	mtime := time.Now().UnixNano()
-	defer cleanupFn()
-
-	chunk, collection, replication, err := pages.fh.f.wfs.saveDataAsChunk(pages.fh.f.fullpath())(reader, pages.fh.f.Name, offset)
-	if err != nil {
-		glog.V(0).Infof("%s saveToStorage [%d,%d): %v", pages.fh.f.fullpath(), offset, offset+size, err)
-		pages.lastErr = err
-		return
-	}
-	chunk.Mtime = mtime
-	pages.collection, pages.replication = collection, replication
-	pages.chunkAddLock.Lock()
-	pages.fh.f.addChunks([]*filer_pb.FileChunk{chunk})
-	pages.fh.entryViewCache = nil
-	glog.V(3).Infof("%s saveToStorage %s [%d,%d)", pages.fh.f.fullpath(), chunk.FileId, offset, offset+size)
-	pages.chunkAddLock.Unlock()
-
-}
-
-func (pages ChunkedDirtyPages) Destroy() {
-	pages.uploadPipeline.Shutdown()
-}
-
-func (pages *ChunkedDirtyPages) LockForRead(startOffset, stopOffset int64) {
-	pages.uploadPipeline.LockForRead(startOffset, stopOffset)
-}
-func (pages *ChunkedDirtyPages) UnlockForRead(startOffset, stopOffset int64) {
-	pages.uploadPipeline.UnlockForRead(startOffset, stopOffset)
-}

weed/filesys/file.go

@@ -1,406 +0,0 @@
-package filesys
-
-import (
-	"context"
-	"os"
-	"sort"
-	"time"
-
-	"github.com/seaweedfs/fuse"
-	"github.com/seaweedfs/fuse/fs"
-
-	"github.com/chrislusf/seaweedfs/weed/filer"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-	"github.com/chrislusf/seaweedfs/weed/util"
-)
-
-const blockSize = 512
-
-var _ = fs.Node(&File{})
-var _ = fs.NodeIdentifier(&File{})
-var _ = fs.NodeOpener(&File{})
-var _ = fs.NodeFsyncer(&File{})
-var _ = fs.NodeSetattrer(&File{})
-var _ = fs.NodeGetxattrer(&File{})
-var _ = fs.NodeSetxattrer(&File{})
-var _ = fs.NodeRemovexattrer(&File{})
-var _ = fs.NodeListxattrer(&File{})
-var _ = fs.NodeForgetter(&File{})
-
-type File struct {
-	Name          string
-	dir           *Dir
-	wfs           *WFS
-	entry         *filer_pb.Entry
-	isOpen        int
-	dirtyMetadata bool
-	id            uint64
-}
-
-func (file *File) fullpath() util.FullPath {
-	return util.NewFullPath(file.dir.FullPath(), file.Name)
-}
-
-func (file *File) Id() uint64 {
-	return file.id
-}
-
-func (file *File) Attr(ctx context.Context, attr *fuse.Attr) (err error) {
-
-	glog.V(4).Infof("file Attr %s, open:%v existing:%v", file.fullpath(), file.isOpen, attr)
-
-	entry, err := file.maybeLoadEntry(ctx)
-	if err != nil {
-		return err
-	}
-
-	if entry == nil {
-		return fuse.ENOENT
-	}
-
-	attr.Inode = file.Id()
-	attr.Valid = time.Second
-	attr.Mode = os.FileMode(entry.Attributes.FileMode)
-	attr.Size = filer.FileSize(entry)
-	if file.isOpen > 0 {
-		attr.Size = entry.Attributes.FileSize
-		glog.V(4).Infof("file Attr %s, open:%v, size: %d", file.fullpath(), file.isOpen, attr.Size)
-	}
-	attr.Crtime = time.Unix(entry.Attributes.Crtime, 0)
-	attr.Ctime = time.Unix(entry.Attributes.Mtime, 0)
-	attr.Mtime = time.Unix(entry.Attributes.Mtime, 0)
-	attr.Gid = entry.Attributes.Gid
-	attr.Uid = entry.Attributes.Uid
-	attr.BlockSize = blockSize
-	attr.Blocks = (attr.Size + blockSize - 1) / blockSize
-	if entry.HardLinkCounter > 0 {
-		attr.Nlink = uint32(entry.HardLinkCounter)
-	}
-
-	return nil
-
-}
-
-func (file *File) Getxattr(ctx context.Context, req *fuse.GetxattrRequest, resp *fuse.GetxattrResponse) error {
-
-	// glog.V(4).Infof("file Getxattr %s", file.fullpath())
-
-	entry, err := file.maybeLoadEntry(ctx)
-	if err != nil {
-		return err
-	}
-
-	return getxattr(entry, req, resp)
-}
-
-func (file *File) Open(ctx context.Context, req *fuse.OpenRequest, resp *fuse.OpenResponse) (fs.Handle, error) {
-
-	glog.V(4).Infof("file %v open %+v", file.fullpath(), req)
-	// resp.Flags |= fuse.OpenDirectIO
-
-	handle := file.wfs.AcquireHandle(file, req.Uid, req.Gid)
-
-	resp.Handle = fuse.HandleID(handle.handle)
-
-	glog.V(4).Infof("%v file open handle id = %d", file.fullpath(), handle.handle)
-
-	return handle, nil
-
-}
-
-func (file *File) Setattr(ctx context.Context, req *fuse.SetattrRequest, resp *fuse.SetattrResponse) error {
-
-	glog.V(4).Infof("%v file setattr %+v mode=%d", file.fullpath(), req, req.Mode)
-
-	entry, err := file.maybeLoadEntry(ctx)
-	if err != nil {
-		return err
-	}
-
-	if req.Valid.Size() {
-
-		glog.V(4).Infof("%v file setattr set size=%v chunks=%d", file.fullpath(), req.Size, len(entry.Chunks))
-		if req.Size < filer.FileSize(entry) {
-			// fmt.Printf("truncate %v \n", fullPath)
-			var chunks []*filer_pb.FileChunk
-			var truncatedChunks []*filer_pb.FileChunk
-			for _, chunk := range entry.Chunks {
-				int64Size := int64(chunk.Size)
-				if chunk.Offset+int64Size > int64(req.Size) {
-					// this chunk is truncated
-					int64Size = int64(req.Size) - chunk.Offset
-					if int64Size > 0 {
-						chunks = append(chunks, chunk)
-						glog.V(4).Infof("truncated chunk %+v from %d to %d\n", chunk.GetFileIdString(), chunk.Size, int64Size)
-						chunk.Size = uint64(int64Size)
-					} else {
-						glog.V(4).Infof("truncated whole chunk %+v\n", chunk.GetFileIdString())
-						truncatedChunks = append(truncatedChunks, chunk)
-					}
-				}
-			}
-			// set the new chunks and reset entry cache
-			entry.Chunks = chunks
-			file.wfs.handlesLock.Lock()
-			existingHandle, found := file.wfs.handles[file.Id()]
-			file.wfs.handlesLock.Unlock()
-			if found {
-				existingHandle.entryViewCache = nil
-			}
-
-		}
-		entry.Attributes.Mtime = time.Now().Unix()
-		entry.Attributes.FileSize = req.Size
-		file.dirtyMetadata = true
-	}
-
-	if req.Valid.Mode() && entry.Attributes.FileMode != uint32(req.Mode) {
-		entry.Attributes.FileMode = uint32(req.Mode)
-		entry.Attributes.Mtime = time.Now().Unix()
-		file.dirtyMetadata = true
-	}
-
-	if req.Valid.Uid() && entry.Attributes.Uid != req.Uid {
-		entry.Attributes.Uid = req.Uid
-		entry.Attributes.Mtime = time.Now().Unix()
-		file.dirtyMetadata = true
-	}
-
-	if req.Valid.Gid() && entry.Attributes.Gid != req.Gid {
-		entry.Attributes.Gid = req.Gid
-		entry.Attributes.Mtime = time.Now().Unix()
-		file.dirtyMetadata = true
-	}
-
-	if req.Valid.Crtime() {
-		entry.Attributes.Crtime = req.Crtime.Unix()
-		entry.Attributes.Mtime = time.Now().Unix()
-		file.dirtyMetadata = true
-	}
-
-	if req.Valid.Mtime() && entry.Attributes.Mtime != req.Mtime.Unix() {
-		entry.Attributes.Mtime = req.Mtime.Unix()
-		file.dirtyMetadata = true
-	}
-
-	if req.Valid.Handle() {
-		// fmt.Printf("file handle => %d\n", req.Handle)
-	}
-
-	if file.isOpen > 0 {
-		return nil
-	}
-
-	if !file.dirtyMetadata {
-		return nil
-	}
-
-	return file.saveEntry(entry)
-
-}
-
-func (file *File) Setxattr(ctx context.Context, req *fuse.SetxattrRequest) error {
-
-	glog.V(4).Infof("file Setxattr %s: %s", file.fullpath(), req.Name)
-
-	entry, err := file.maybeLoadEntry(ctx)
-	if err != nil {
-		return err
-	}
-
-	if err := setxattr(entry, req); err != nil {
-		return err
-	}
-	file.dirtyMetadata = true
-
-	if file.isOpen > 0 {
-		return nil
-	}
-
-	return file.saveEntry(entry)
-
-}
-
-func (file *File) Removexattr(ctx context.Context, req *fuse.RemovexattrRequest) error {
-
-	glog.V(4).Infof("file Removexattr %s: %s", file.fullpath(), req.Name)
-
-	entry, err := file.maybeLoadEntry(ctx)
-	if err != nil {
-		return err
-	}
-
-	if err := removexattr(entry, req); err != nil {
-		return err
-	}
-	file.dirtyMetadata = true
-
-	if file.isOpen > 0 {
-		return nil
-	}
-
-	return file.saveEntry(entry)
-
-}
-
-func (file *File) Listxattr(ctx context.Context, req *fuse.ListxattrRequest, resp *fuse.ListxattrResponse) error {
-
-	glog.V(4).Infof("file Listxattr %s", file.fullpath())
-
-	entry, err := file.maybeLoadEntry(ctx)
-	if err != nil {
-		return err
-	}
-
-	if err := listxattr(entry, req, resp); err != nil {
-		return err
-	}
-
-	return nil
-
-}
-
-func (file *File) Fsync(ctx context.Context, req *fuse.FsyncRequest) error {
-
-	// write the file chunks to the filerGrpcAddress
-	glog.V(4).Infof("%s/%s fsync file %+v", file.dir.FullPath(), file.Name, req)
-
-	return file.wfs.Fsync(file, req.Header)
-
-}
-
-func (file *File) Forget() {
-	t := util.NewFullPath(file.dir.FullPath(), file.Name)
-	glog.V(4).Infof("Forget file %s", t)
-	file.wfs.ReleaseHandle(t, fuse.HandleID(t.AsInode(file.entry.FileMode())))
-
-}
-
-func (file *File) maybeLoadEntry(ctx context.Context) (entry *filer_pb.Entry, err error) {
-
-	file.wfs.handlesLock.Lock()
-	handle, found := file.wfs.handles[file.Id()]
-	file.wfs.handlesLock.Unlock()
-	entry = file.entry
-	if found {
-		// glog.V(4).Infof("maybeLoadEntry found opened file %s/%s", file.dir.FullPath(), file.Name)
-		entry = handle.f.entry
-	}
-
-	if entry != nil {
-		if len(entry.HardLinkId) == 0 {
-			// only always reload hard link
-			return entry, nil
-		}
-	}
-	entry, err = file.wfs.maybeLoadEntry(file.dir.FullPath(), file.Name)
-	if err != nil {
-		glog.V(3).Infof("maybeLoadEntry file %s/%s: %v", file.dir.FullPath(), file.Name, err)
-		return entry, err
-	}
-	if entry != nil {
-		// file.entry = entry
-	} else {
-		glog.Warningf("maybeLoadEntry not found entry %s/%s: %v", file.dir.FullPath(), file.Name, err)
-	}
-	return entry, nil
-}
-
-func lessThan(a, b *filer_pb.FileChunk) bool {
-	if a.Mtime == b.Mtime {
-		return a.Fid.FileKey < b.Fid.FileKey
-	}
-	return a.Mtime < b.Mtime
-}
-
-func (file *File) addChunks(chunks []*filer_pb.FileChunk) {
-
-	// find the earliest incoming chunk
-	newChunks := chunks
-	earliestChunk := newChunks[0]
-	for i := 1; i < len(newChunks); i++ {
-		if lessThan(earliestChunk, newChunks[i]) {
-			earliestChunk = newChunks[i]
-		}
-	}
-
-	entry := file.getEntry()
-	if entry == nil {
-		return
-	}
-
-	// pick out-of-order chunks from existing chunks
-	for _, chunk := range entry.Chunks {
-		if lessThan(earliestChunk, chunk) {
-			chunks = append(chunks, chunk)
-		}
-	}
-
-	// sort incoming chunks
-	sort.Slice(chunks, func(i, j int) bool {
-		return lessThan(chunks[i], chunks[j])
-	})
-
-	glog.V(4).Infof("%s existing %d chunks adds %d more", file.fullpath(), len(entry.Chunks), len(chunks))
-
-	entry.Chunks = append(entry.Chunks, newChunks...)
-}
-
-func (file *File) saveEntry(entry *filer_pb.Entry) error {
-	return file.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
-
-		file.wfs.mapPbIdFromLocalToFiler(entry)
-		defer file.wfs.mapPbIdFromFilerToLocal(entry)
-
-		request := &filer_pb.CreateEntryRequest{
-			Directory:  file.dir.FullPath(),
-			Entry:      entry,
-			Signatures: []int32{file.wfs.signature},
-		}
-
-		glog.V(4).Infof("save file entry: %v", request)
-		_, err := client.CreateEntry(context.Background(), request)
-		if err != nil {
-			glog.Errorf("UpdateEntry file %s/%s: %v", file.dir.FullPath(), file.Name, err)
-			return fuse.EIO
-		}
-
-		file.wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry))
-
-		file.dirtyMetadata = false
-
-		return nil
-	})
-}
-
-func (file *File) getEntry() *filer_pb.Entry {
-	return file.entry
-}
-
-func (file *File) downloadRemoteEntry(entry *filer_pb.Entry) (*filer_pb.Entry, error) {
-	err := file.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
-
-		request := &filer_pb.CacheRemoteObjectToLocalClusterRequest{
-			Directory: file.dir.FullPath(),
-			Name:      entry.Name,
-		}
-
-		glog.V(4).Infof("download entry: %v", request)
-		resp, err := client.CacheRemoteObjectToLocalCluster(context.Background(), request)
-		if err != nil {
-			glog.Errorf("CacheRemoteObjectToLocalCluster file %s/%s: %v", file.dir.FullPath(), file.Name, err)
-			return fuse.EIO
-		}
-
-		entry = resp.Entry
-
-		file.wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, resp.Entry))
-
-		file.dirtyMetadata = false
-
-		return nil
-	})
-
-	return entry, err
-}

weed/filesys/filehandle.go

@@ -1,347 +0,0 @@
-package filesys
-
-import (
-	"context"
-	"fmt"
-	"io"
-	"math"
-	"net/http"
-	"os"
-	"sync"
-	"time"
-
-	"github.com/seaweedfs/fuse"
-	"github.com/seaweedfs/fuse/fs"
-
-	"github.com/chrislusf/seaweedfs/weed/filer"
-	"github.com/chrislusf/seaweedfs/weed/glog"
-	"github.com/chrislusf/seaweedfs/weed/pb/filer_pb"
-)
-
-type FileHandle struct {
-	// cache file has been written to
-	dirtyPages     *PageWriter
-	entryViewCache []filer.VisibleInterval
-	reader         io.ReaderAt
-	contentType    string
-	handle         uint64
-	sync.Mutex
-
-	f         *File
-	NodeId    fuse.NodeID // file or directory the request is about
-	Uid       uint32      // user ID of process making request
-	Gid       uint32      // group ID of process making request
-	isDeleted bool
-}
-
-func newFileHandle(file *File, uid, gid uint32) *FileHandle {
-	fh := &FileHandle{
-		f:   file,
-		Uid: uid,
-		Gid: gid,
-	}
-	// dirtyPages: newContinuousDirtyPages(file, writeOnly),
-	fh.dirtyPages = newPageWriter(fh, file.wfs.option.ChunkSizeLimit)
-	entry := fh.f.getEntry()
-	if entry != nil {
-		entry.Attributes.FileSize = filer.FileSize(entry)
-	}
-
-	return fh
-}
-
-var _ = fs.Handle(&FileHandle{})
-
-// var _ = fs.HandleReadAller(&FileHandle{})
-var _ = fs.HandleReader(&FileHandle{})
-var _ = fs.HandleFlusher(&FileHandle{})
-var _ = fs.HandleWriter(&FileHandle{})
-var _ = fs.HandleReleaser(&FileHandle{})
-
-func (fh *FileHandle) Read(ctx context.Context, req *fuse.ReadRequest, resp *fuse.ReadResponse) error {
-
-	fh.Lock()
-	defer fh.Unlock()
-
-	glog.V(4).Infof("%s read fh %d: [%d,%d) size %d resp.Data cap=%d", fh.f.fullpath(), fh.handle, req.Offset, req.Offset+int64(req.Size), req.Size, cap(resp.Data))
-
-	if req.Size <= 0 {
-		return nil
-	}
-
-	buff := resp.Data[:cap(resp.Data)]
-	if req.Size > cap(resp.Data) {
-		// should not happen
-		buff = make([]byte, req.Size)
-	}
-
-	fh.lockForRead(req.Offset, len(buff))
-	defer fh.unlockForRead(req.Offset, len(buff))
-	totalRead, err := fh.readFromChunks(buff, req.Offset)
-	if err == nil || err == io.EOF {
-		maxStop := fh.readFromDirtyPages(buff, req.Offset)
-		totalRead = max(maxStop-req.Offset, totalRead)
-	}
-
-	if err == io.EOF {
-		err = nil
-	}
-
-	if err != nil {
-		glog.Warningf("file handle read %s %d: %v", fh.f.fullpath(), totalRead, err)
-		return fuse.EIO
-	}
-
-	if totalRead > int64(len(buff)) {
-		glog.Warningf("%s FileHandle Read %d: [%d,%d) size %d totalRead %d", fh.f.fullpath(), fh.handle, req.Offset, req.Offset+int64(req.Size), req.Size, totalRead)
-		totalRead = min(int64(len(buff)), totalRead)
-	}
-	if err == nil {
-		resp.Data = buff[:totalRead]
-	}
-
-	return err
-}
-
-func (fh *FileHandle) lockForRead(startOffset int64, size int) {
-	fh.dirtyPages.LockForRead(startOffset, startOffset+int64(size))
-}
-func (fh *FileHandle) unlockForRead(startOffset int64, size int) {
-	fh.dirtyPages.UnlockForRead(startOffset, startOffset+int64(size))
-}
-
-func (fh *FileHandle) readFromDirtyPages(buff []byte, startOffset int64) (maxStop int64) {
-	maxStop = fh.dirtyPages.ReadDirtyDataAt(buff, startOffset)
-	return
-}
-
-func (fh *FileHandle) readFromChunks(buff []byte, offset int64) (int64, error) {
-
-	entry := fh.f.getEntry()
-	if entry == nil {
-		return 0, io.EOF
-	}
-
-	if entry.IsInRemoteOnly() {
-		glog.V(4).Infof("download remote entry %s", fh.f.fullpath())
-		newEntry, err := fh.f.downloadRemoteEntry(entry)
-		if err != nil {
-			glog.V(1).Infof("download remote entry %s: %v", fh.f.fullpath(), err)
-			return 0, err
-		}
-		entry = newEntry
-	}
-
-	fileSize := int64(filer.FileSize(entry))
-	fileFullPath := fh.f.fullpath()
-
-	if fileSize == 0 {
-		glog.V(1).Infof("empty fh %v", fileFullPath)
-		return 0, io.EOF
-	}
-
-	if offset+int64(len(buff)) <= int64(len(entry.Content)) {
-		totalRead := copy(buff, entry.Content[offset:])
-		glog.V(4).Infof("file handle read cached %s [%d,%d] %d", fileFullPath, offset, offset+int64(totalRead), totalRead)
-		return int64(totalRead), nil
-	}
-
-	var chunkResolveErr error
-	if fh.entryViewCache == nil {
-		fh.entryViewCache, chunkResolveErr = filer.NonOverlappingVisibleIntervals(fh.f.wfs.LookupFn(), entry.Chunks, 0, math.MaxInt64)
-		if chunkResolveErr != nil {
-			return 0, fmt.Errorf("fail to resolve chunk manifest: %v", chunkResolveErr)
-		}
-		fh.reader = nil
-	}
-
-	reader := fh.reader
-	if reader == nil {
-		chunkViews := filer.ViewFromVisibleIntervals(fh.entryViewCache, 0, math.MaxInt64)
-		glog.V(4).Infof("file handle read %s [%d,%d) from %d views", fileFullPath, offset, offset+int64(len(buff)), len(chunkViews))
-		for _, chunkView := range chunkViews {
-			glog.V(4).Infof("  read %s [%d,%d) from chunk %+v", fileFullPath, chunkView.LogicOffset, chunkView.LogicOffset+int64(chunkView.Size), chunkView.FileId)
-		}
-		reader = filer.NewChunkReaderAtFromClient(fh.f.wfs.LookupFn(), chunkViews, fh.f.wfs.chunkCache, fileSize)
-	}
-	fh.reader = reader
-
-	totalRead, err := reader.ReadAt(buff, offset)
-
-	if err != nil && err != io.EOF {
-		glog.Errorf("file handle read %s: %v", fileFullPath, err)
-	}
-
-	glog.V(4).Infof("file handle read %s [%d,%d] %d : %v", fileFullPath, offset, offset+int64(totalRead), totalRead, err)
-
-	return int64(totalRead), err
-}
-
-// Write to the file handle
-func (fh *FileHandle) Write(ctx context.Context, req *fuse.WriteRequest, resp *fuse.WriteResponse) error {
-
-	fh.dirtyPages.writerPattern.MonitorWriteAt(req.Offset, len(req.Data))
-
-	fh.Lock()
-	defer fh.Unlock()
-
-	// write the request to volume servers
-	data := req.Data
-	if len(data) <= 512 && req.Offset == 0 {
-		// fuse message cacheable size
-		data = make([]byte, len(req.Data))
-		copy(data, req.Data)
-	}
-
-	entry := fh.f.getEntry()
-	if entry == nil {
-		return fuse.EIO
-	}
-
-	entry.Content = nil
-	entry.Attributes.FileSize = uint64(max(req.Offset+int64(len(data)), int64(entry.Attributes.FileSize)))
-	// glog.V(4).Infof("%v write [%d,%d) %d", fh.f.fullpath(), req.Offset, req.Offset+int64(len(req.Data)), len(req.Data))
-
-	fh.dirtyPages.AddPage(req.Offset, data)
-
-	resp.Size = len(data)
-
-	if req.Offset == 0 {
-		// detect mime type
-		fh.contentType = http.DetectContentType(data)
-		fh.f.dirtyMetadata = true
-	}
-
-	fh.f.dirtyMetadata = true
-
-	return nil
-}
-
-func (fh *FileHandle) Release(ctx context.Context, req *fuse.ReleaseRequest) error {
-
-	glog.V(4).Infof("Release %v fh %d open=%d", fh.f.fullpath(), fh.handle, fh.f.isOpen)
-
-	fh.f.wfs.handlesLock.Lock()
-	fh.f.isOpen--
-	fh.f.wfs.handlesLock.Unlock()
-
-	if fh.f.isOpen <= 0 {
-		fh.f.entry = nil
-		fh.entryViewCache = nil
-		fh.reader = nil
-
-		fh.f.wfs.ReleaseHandle(fh.f.fullpath(), fuse.HandleID(fh.handle))
-		fh.dirtyPages.Destroy()
-	}
-
-	if fh.f.isOpen < 0 {
-		glog.V(0).Infof("Release reset %s open count %d => %d", fh.f.Name, fh.f.isOpen, 0)
-		fh.f.isOpen = 0
-		return nil
-	}
-
-	return nil
-}
-
-func (fh *FileHandle) Flush(ctx context.Context, req *fuse.FlushRequest) error {
-
-	glog.V(4).Infof("Flush %v fh %d", fh.f.fullpath(), fh.handle)
-
-	if fh.isDeleted {
-		glog.V(4).Infof("Flush %v fh %d skip deleted", fh.f.fullpath(), fh.handle)
-		return nil
-	}
-
-	fh.Lock()
-	defer fh.Unlock()
-
-	if err := fh.doFlush(ctx, req.Header); err != nil {
-		glog.Errorf("Flush doFlush %s: %v", fh.f.Name, err)
-		return err
-	}
-
-	return nil
-}
-
-func (fh *FileHandle) doFlush(ctx context.Context, header fuse.Header) error {
-	// flush works at fh level
-	// send the data to the OS
-	glog.V(4).Infof("doFlush %s fh %d", fh.f.fullpath(), fh.handle)
-
-	if err := fh.dirtyPages.FlushData(); err != nil {
-		glog.Errorf("%v doFlush: %v", fh.f.fullpath(), err)
-		return fuse.EIO
-	}
-
-	if !fh.f.dirtyMetadata {
-		return nil
-	}
-
-	err := fh.f.wfs.WithFilerClient(false, func(client filer_pb.SeaweedFilerClient) error {
-
-		entry := fh.f.getEntry()
-		if entry == nil {
-			return nil
-		}
-
-		if entry.Attributes != nil {
-			entry.Attributes.Mime = fh.contentType
-			if entry.Attributes.Uid == 0 {
-				entry.Attributes.Uid = header.Uid
-			}
-			if entry.Attributes.Gid == 0 {
-				entry.Attributes.Gid = header.Gid
-			}
-			if entry.Attributes.Crtime == 0 {
-				entry.Attributes.Crtime = time.Now().Unix()
-			}
-			entry.Attributes.Mtime = time.Now().Unix()
-			entry.Attributes.FileMode = uint32(os.FileMode(entry.Attributes.FileMode) &^ fh.f.wfs.option.Umask)
-			entry.Attributes.Collection, entry.Attributes.Replication = fh.dirtyPages.GetStorageOptions()
-		}
-
-		request := &filer_pb.CreateEntryRequest{
-			Directory:  fh.f.dir.FullPath(),
-			Entry:      entry,
-			Signatures: []int32{fh.f.wfs.signature},
-		}
-
-		glog.V(4).Infof("%s set chunks: %v", fh.f.fullpath(), len(entry.Chunks))
-		for i, chunk := range entry.Chunks {
-			glog.V(4).Infof("%s chunks %d: %v [%d,%d)", fh.f.fullpath(), i, chunk.GetFileIdString(), chunk.Offset, chunk.Offset+int64(chunk.Size))
-		}
-
-		manifestChunks, nonManifestChunks := filer.SeparateManifestChunks(entry.Chunks)
-
-		chunks, _ := filer.CompactFileChunks(fh.f.wfs.LookupFn(), nonManifestChunks)
-		chunks, manifestErr := filer.MaybeManifestize(fh.f.wfs.saveDataAsChunk(fh.f.fullpath()), chunks)
-		if manifestErr != nil {
-			// not good, but should be ok
-			glog.V(0).Infof("MaybeManifestize: %v", manifestErr)
-		}
-		entry.Chunks = append(chunks, manifestChunks...)
-
-		fh.f.wfs.mapPbIdFromLocalToFiler(request.Entry)
-		defer fh.f.wfs.mapPbIdFromFilerToLocal(request.Entry)
-
-		if err := filer_pb.CreateEntry(client, request); err != nil {
-			glog.Errorf("fh flush create %s: %v", fh.f.fullpath(), err)
-			return fmt.Errorf("fh flush create %s: %v", fh.f.fullpath(), err)
-		}
-
-		fh.f.wfs.metaCache.InsertEntry(context.Background(), filer.FromPbEntry(request.Directory, request.Entry))
-
-		return nil
-	})
-
-	if err == nil {
-		fh.f.dirtyMetadata = false
-	}
-
-	if err != nil {
-		glog.Errorf("%v fh %d flush: %v", fh.f.fullpath(), fh.handle, err)
-		return fuse.EIO
-	}
-
-	return nil
-}