improve docker build

.github/workflows/e2e.yml

@@ -32,14 +32,54 @@ jobs:
     - name: Check out code into the Go module directory
       uses: actions/checkout@08c6903cd8c0fde910a37f88322edcfb5dd907a8 # v2
 
+    - name: Set up Docker Buildx
+      uses: docker/setup-buildx-action@v3
+      
+    - name: Cache Docker layers
+      uses: actions/cache@v4
+      with:
+        path: /tmp/.buildx-cache
+        key: ${{ runner.os }}-buildx-e2e-${{ github.sha }}
+        restore-keys: |
+          ${{ runner.os }}-buildx-e2e-
+
     - name: Install dependencies
       run: |
-        sudo apt-get update
-        sudo apt-get install -y fuse
+        # Use faster mirrors and install with timeout
+        echo "deb http://azure.archive.ubuntu.com/ubuntu/ $(lsb_release -cs) main restricted universe multiverse" | sudo tee /etc/apt/sources.list
+        echo "deb http://azure.archive.ubuntu.com/ubuntu/ $(lsb_release -cs)-updates main restricted universe multiverse" | sudo tee -a /etc/apt/sources.list
+        
+        sudo apt-get update --fix-missing
+        sudo DEBIAN_FRONTEND=noninteractive apt-get install -y --no-install-recommends fuse
+        
+        # Verify FUSE installation
+        echo "FUSE version: $(fusermount --version 2>&1 || echo 'fusermount not found')"
+        echo "FUSE device: $(ls -la /dev/fuse 2>&1 || echo '/dev/fuse not found')"
 
     - name: Start SeaweedFS
-      timeout-minutes: 5
-      run: make build_e2e && docker compose -f ./compose/e2e-mount.yml up --wait
+      timeout-minutes: 10
+      run: |
+        # Enable Docker buildkit for better caching
+        export DOCKER_BUILDKIT=1
+        export COMPOSE_DOCKER_CLI_BUILD=1
+        
+        # Build with retry logic
+        for i in {1..3}; do
+          echo "Build attempt $i/3"
+          if make build_e2e; then
+            echo "Build successful on attempt $i"
+            break
+          elif [ $i -eq 3 ]; then
+            echo "Build failed after 3 attempts"
+            exit 1
+          else
+            echo "Build attempt $i failed, retrying in 30 seconds..."
+            sleep 30
+          fi
+        done
+        
+        # Start services with wait
+        docker compose -f ./compose/e2e-mount.yml up --wait
 
     - name: Run FIO 4k
       timeout-minutes: 15

docker/Dockerfile.e2e

@@ -2,7 +2,18 @@ FROM ubuntu:22.04
 
 LABEL author="Chris Lu"
 
-RUN apt-get update && apt-get install -y curl fio fuse
+# Use faster mirrors and optimize package installation
+RUN apt-get update && \
+    DEBIAN_FRONTEND=noninteractive apt-get install -y \
+    --no-install-recommends \
+    --no-install-suggests \
+    curl \
+    fio \
+    fuse \
+    && apt-get clean \
+    && rm -rf /var/lib/apt/lists/* \
+    && rm -rf /tmp/* \
+    && rm -rf /var/tmp/*
 RUN mkdir -p /etc/seaweedfs /data/filerldb2
 
 COPY ./weed /usr/bin/

docker/Makefile

@@ -20,7 +20,14 @@ build: binary
 	docker build --no-cache -t chrislusf/seaweedfs:local -f Dockerfile.local .
 
 build_e2e: binary_race
-	docker build --no-cache -t chrislusf/seaweedfs:e2e -f Dockerfile.e2e .
+	DOCKER_BUILDKIT=1 docker build \
+		--cache-from=type=local,src=/tmp/.buildx-cache \
+		--cache-to=type=local,dest=/tmp/.buildx-cache-new,mode=max \
+		-t chrislusf/seaweedfs:e2e \
+		-f Dockerfile.e2e .
+	# Move cache to avoid growing cache size
+	rm -rf /tmp/.buildx-cache || true
+	mv /tmp/.buildx-cache-new /tmp/.buildx-cache || true
 
 go_build: # make go_build tags=elastic,ydb,gocdk,hdfs,5BytesOffset,tarantool
 	docker build --build-arg TAGS=$(tags) --no-cache -t chrislusf/seaweedfs:go_build -f Dockerfile.go_build .

docker/compose/e2e-mount.yml

@@ -6,16 +6,20 @@ services:
     command: "-v=4 master -ip=master -ip.bind=0.0.0.0 -raftBootstrap"
     healthcheck:
       test: [ "CMD", "curl", "--fail", "-I", "http://localhost:9333/cluster/healthz" ]
-      interval: 1s
-      timeout: 60s
+      interval: 2s
+      timeout: 10s
+      retries: 30
+      start_period: 10s
 
   volume:
     image: chrislusf/seaweedfs:e2e
     command: "-v=4 volume -mserver=master:9333 -ip=volume -ip.bind=0.0.0.0 -preStopSeconds=1"
     healthcheck:
       test: [ "CMD", "curl", "--fail", "-I", "http://localhost:8080/healthz" ]
-      interval: 1s
-      timeout: 30s
+      interval: 2s
+      timeout: 10s
+      retries: 15
+      start_period: 5s
     depends_on:
       master:
         condition: service_healthy
@@ -25,8 +29,10 @@ services:
     command: "-v=4 filer -master=master:9333 -ip=filer -ip.bind=0.0.0.0"
     healthcheck:
       test: [ "CMD", "curl", "--fail", "-I", "http://localhost:8888" ]
-      interval: 1s
-      timeout: 30s
+      interval: 2s
+      timeout: 10s
+      retries: 15
+      start_period: 5s
     depends_on:
       volume:
         condition: service_healthy
@@ -46,8 +52,10 @@ services:
           memory: 4096m
     healthcheck:
       test: [ "CMD", "mountpoint", "-q", "--", "/mnt/seaweedfs" ]
-      interval: 1s
-      timeout: 30s
+      interval: 2s
+      timeout: 10s
+      retries: 15
+      start_period: 10s
     depends_on:
       filer:
         condition: service_healthy