太长不看：重启大法好，sudo reboot
先说个题外话，dockerd是用sudo启动的话，docker-compose也要用sudo。
有一次排查问题，我多次重启docker
sudo systemctl restart docker
并且删除了/run/containerd/containerd.sock和/var/run/containerd/containerd.sock两个socket文件
1
2
| sudo ls -l /run/containerd/containerd.sock
srw-rw----. 1 root root 0 Oct 10 13:59 /run/containerd/containerd.sock
|
然后docker就起不来了 >.<
那就排查问题，先看看sudo systemctl status docker -l和sudo journalctl -xe
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
| ● docker.service - Docker Application Container Engine
Loaded: loaded (/usr/lib/systemd/system/docker.service; disabled; vendor preset: disabled)
Active: activating (start) since Thu 2019-10-10 13:48:27 CST; 14s ago
Docs: https://docs.docker.com
Main PID: 25992 (dockerd)
Tasks: 12
Memory: 39.7M
CGroup: /system.slice/docker.service
└─25992 /usr/bin/dockerd -H fd:// --containerd=/run/containerd/containerd.sock
Oct 10 13:48:27 node-1.local systemd[1]: Starting Docker Application Container Engine...
Oct 10 13:48:27 node-1.local dockerd[25992]: time="2019-10-10T13:48:27.700450359+08:00" level=info msg="Starting up"
Oct 10 13:48:27 node-1.local dockerd[25992]: time="2019-10-10T13:48:27.701786486+08:00" level=info msg="parsed scheme: \"unix\"" module=grpc
Oct 10 13:48:27 node-1.local dockerd[25992]: time="2019-10-10T13:48:27.701818831+08:00" level=info msg="scheme \"unix\" not registered, fallback to default scheme" module=grpc
Oct 10 13:48:27 node-1.local dockerd[25992]: time="2019-10-10T13:48:27.701856021+08:00" level=info msg="ccResolverWrapper: sending update to cc: {[{unix:///run/containerd/containerd.sock 0 <nil>}] <nil>}" module=grpc
Oct 10 13:48:27 node-1.local dockerd[25992]: time="2019-10-10T13:48:27.701880147+08:00" level=info msg="ClientConn switching balancer to \"pick_first\"" module=grpc
|
看不出问题，那就用debug模式启动docker看看
sudo /usr/bin/dockerd --containerd=/run/containerd/containerd.sock
1
2
3
4
5
6
7
8
9
10
| INFO[2019-10-10T14:35:29.260193095+08:00] Starting up
DEBU[2019-10-10T14:35:29.260691124+08:00] Listener created for HTTP on unix (/var/run/docker.sock)
DEBU[2019-10-10T14:35:29.261151949+08:00] Golang's threads limit set to 57240
INFO[2019-10-10T14:35:29.261477776+08:00] parsed scheme: "unix" module=grpc
INFO[2019-10-10T14:35:29.261496182+08:00] scheme "unix" not registered, fallback to default scheme module=grpc
INFO[2019-10-10T14:35:29.261518855+08:00] ccResolverWrapper: sending update to cc: {[{unix:///run/containerd/containerd.sock 0 <nil>}] <nil>} module=grpc
INFO[2019-10-10T14:35:29.261529702+08:00] ClientConn switching balancer to "pick_first" module=grpc
WARN[2019-10-10T14:35:29.262168877+08:00] grpc: addrConn.createTransport failed to connect to {unix:///run/containerd/containerd.sock 0 <nil>}. Err :connection error: desc = "transport: Error while dialing dial unix /run/containerd/containerd.sock: connect: connection refused". Reconnecting... module=grpc
DEBU[2019-10-10T14:35:29.262315604+08:00] Cleaning up old mountid : start.
failed to start daemon: failed to dial "/run/containerd/containerd.sock": all SubConns are in TransientFailure, latest connection error: connection error: desc = "transport: Error while dialing dial unix /run/containerd/containerd.sock: connect: connection refused": unavailable
|
找到问题了，是连接不到/run/containerd/containerd.sock，那被我删了嘛，docker自己不会新建嘛？那我帮它创建好了。
创建socket文件的方式有mksock、nc或者用python
sudo mksocket /run/containerd/containerd.sock
sudo nc -lU /run/containerd/containerd.sock
python -c "import socket as s; sock = s.socket(s.AF_UNIX); sock.bind('/run/containerd/containerd.sock')"
继续debug模式启动docker，可是得到了上述同样的报错
最后我把docker卸载了重装也不行，无奈之下只好….
重启大法
sudo reboot
重启docker服务，搞定。