Docker容器强制删除及文件系统修复的完整指南

# 1. 强制删除问题容器
docker rm -f c9ca40be974d

# 2. 尝试系统级清理
docker system prune -af --volumes

# 3. 重启Docker服务
sudo systemctl restart docker

# 1. 停止Docker服务
sudo systemctl stop docker

# 2. 卸载相关存储（注意确认挂载点）
umount /data/docker/storage

# 3. 执行文件系统修复（按类型选择）
# XFS文件系统：
sudo xfs_repair /dev/sdXX
# EXT4文件系统：
sudo fsck -y /dev/sdXX

# 4. 手动清理残留目录
sudo rm -rf /data/docker/storage/containers/c9ca40be974d0aebc42ac9471bdbcd2752bf86c107f07361d6f578d02d98eae1

# 5. 重挂载并重启服务
mount /data/docker/storage
sudo systemctl start docker

# 1. 检查容器残留
docker ps -a | grep c9ca40be974d

# 2. 验证存储驱动状态
docker info | grep "Storage Driver"

# 3. 重建服务栈
docker-compose up -d --force-recreate

# 检查磁盘健康
smartctl -a /dev/sdX

# 查看文件系统错误日志
xfs_info /data/docker/storage  # XFS系统
dumpe2fs /dev/sdXX | grep -i error  # EXT4系统

# 容器层深度检测
docker diff c9ca40be974d

# 使用容器运行时接口(CRI)直接删除
sudo crictl rm -f c9ca40be974d0aebc42ac9471bdbcd2752bf86c107f07361d6f578d02d98eae1

# 通过libcontainer直接移除
sudo nsenter --mount=/proc/$(docker inspect c9ca40be974d | jq .State.Pid)/ns/mnt 
rm -rf /data/docker/storage/containers/<full_id>

# /etc/docker/daemon.json
{
  "storage-driver": "overlay2",
+ "storage-opts": [
+   "xfs_nospace_max_retries=10",
+   "auto-repair=true"
+ ],
+ "data-root": "/opt/docker"  # 独立存储分区
}

# 添加每日自动巡检
echo "0 3 * * * root xfs_scrub -v /data/docker/storage && docker system prune -f" | sudo tee /etc/cron.d/docker-fs-maintain

mkfs.xfs -n ftype=1 /dev/sdXX  # 确保支持d_type
mount -o pquota /dev/sdXX /data/docker/storage

# docker-compose.yml
services:
  OpIsosMD_OB:
    restart: unless-stopped
    deploy:
      resources:
        limits:
          memory: 2g
           healthcheck:
           test: ["CMD-SHELL", "check_status"]
           interval: 30s

   # /etc/sysctl.conf
   fs.file-max = 10000000
   fs.inotify.max_user_watches = 1048576
   vm.swappiness = 10

graph LR
A[报错出现] --> B{错误类型判断}
B -- Structure needs cleaning --> C[停止Docker]
B -- Device busy --> D[fuser/kill占用进程]
C --> E[umount存储卷]
E --> F[xfs_repair/fsck]
F --> G[文件系统扫描]
G -- 成功 --> H[重建容器]
G -- 失败 --> I[磁盘坏道检测]
I -- 硬件故障 --> J[数据迁移]
I -- 逻辑损坏 --> K[数据重建]