docker空間爆滿導致的進入容器失敗的解決方案

由於問題發生的比較突然,業務催的比較急,所以沒來得及截圖案發現場,所以現場截圖為後期正常的環境。

周一上班,照例進入服務器,進入docker容器

>>> docker exec -i -t xxx /bin/bash
Docker error : no space left on device

沒空間瞭?

立馬想到查看系統的空間

>>> df -h
Filesystem Size Used Avail Use% Mounted on
devtmpfs 3.9G 0 3.9G 0% /dev
tmpfs  3.9G 0 3.9G 0% /dev/shm
tmpfs  3.9G 984K 3.9G 1% /run
tmpfs  3.9G 0 3.9G 0% /sys/fs/cgroup
/dev/vda1 50G 50G 0G 100% /
overlay  50G 5.9G 41G 13% /var/lib/docker/overlay2/4d0941b78fa413f3b77111735e06045b41351748bcea7964205bcfbf9d4ec0b6/merged
overlay  50G 5.9G 41G 13% /var/lib/docker/overlay2/54a7a7b66d9c12d7e42158d177a6b67321f7da1f223b301e838e9bc109a2bda4/merged
shm  64M 0 64M 0% /var/lib/docker/containers/712f65baea85d898f6c948e7149f84f2f5eaf3b1934540603f32ab278f9acec4/mounts/shm
shm  64M 0 64M 0% /var/lib/docker/containers/a5fa673086c8f46ba98fc4425b353ed2e12de3277a5fe2dc5b8b7affa63b4518/mounts/shm
overlay  50G 5.9G 41G 13% /var/lib/docker/overlay2/7fbbc8a29119a1eaa1f212c50b75405a1f16fd68e3ae3949cc0c963d0727a9ab/merged
shm  64M 0 64M 0% /var/lib/docker/containers/011a83deceacecbacb4ef7eb06eb5b812babf9e83914a4fb33d4925cc1ad375b/mounts/shm
tmpfs  783M 0 783M 0% /run/user/0

原來是根目錄滿瞭

進一步查找

>>> cd /
>>> du -h --max-depth=1
984K ./run
16K ./opt
13M ./root
4.0K ./media
du: cannot access ‘./proc/4382/task/4382/fd/4': No such file or directory
du: cannot access ‘./proc/4382/task/4382/fdinfo/4': No such file or directory
du: cannot access ‘./proc/4382/fd/3': No such file or directory
du: cannot access ‘./proc/4382/fdinfo/3': No such file or directory
0 ./proc
204M ./boot
12K ./redis
39M ./etc
16K ./lost+found
4.0K ./srv
0 ./sys
47G ./var
2.8G ./usr
4.0K ./mnt
36K ./tmp
0 ./dev
4.0K ./home
50G .
>>> cd var
>>> du -h --max-depth=1
116M ./cache
8.0K ./empty
4.0K ./games
4.0K ./opt
24K ./db
46G ./lib
4.0K ./gopher
4.0K ./adm
4.0K ./crash
12K ./kerberos
4.0K ./preserve
4.0K ./nis
16K ./tmp
4.0K ./yp
4.0K ./local
104K ./spool
374M ./log
47G .
>>> cd lib
>>> du -h --max-depth=1
76K ./systemd
24K ./NetworkManager
4.0K ./tuned
4.0K ./games
248K ./cloud
215M ./rpm
8.0K ./plymouth
46G ./docker
248K ./containerd
4.0K ./dbus
4.0K ./initramfs
4.0K ./os-prober
8.0K ./rsyslog
24K ./alternatives
8.0K ./authconfig
12K ./stateless
4.0K ./misc
4.0K ./ntp
8.0K ./dhclient
4.0K ./selinux
8.0K ./chrony
4.0K ./rpm-state
12M ./yum
8.0K ./postfix
28K ./polkit-1
4.0K ./machines
8.0K ./logrotate
46G .
>>> cd docker
>>> du -h --max-depth=1
84K ./network
108K ./buildkit
4.0K ./trust
4.0K ./runtimes
5.7M ./image
42G ./volumes
24K ./plugins
4.0K ./tmp
20K ./builder
180K ./containers
3.3G ./overlay2
4.0K ./swarm
46G .
>>> cd volume
>>> du -h --max-depth=1
172K ./kudu_to_jdy_kudu_to_jdy
42G ./jdy_extensions_logs
748K ./bot_etl_bot_etl
42G .

原來是jdy_extensions_logs這個volume爆滿導致的,後面就是找代碼bug的過程瞭,略

後記:

其實docker提供瞭很多命令來對容器鏡像進行管理,其中的docker system df就是一個比較有用的命令

# 顯示docker文件系統使用情況
>>> docker system df
TYPE  TOTAL ACTIVE SIZE RECLAIMABLE
Images  7  3  2.146GB 693.2MB (32%)
Containers 3  3  127.5kB 0B (0%)
Local Volumes 3  3  1.511MB 0B (0%)
Build Cache 0  0  0B 0B
# 顯示空間使用情況
>>> docker system df -v
Images space usage:
REPOSITORY TAG IMAGE ID CREATED SIZE SHARED SIZE UNIQUE SIZE CONTAINERS
jdy_extension 1.7 e3be3be9664a 15 hours ago 627.6MB 73.86MB 553.7MB 1
<none>  <none> 4b5825747ae9 19 hours ago 74.02MB 73.86MB 158.8kB 0
redis  6.2.1 f877e80bb9ef 2 weeks ago 105.3MB 0B  105.3MB 0
kudu_to_jdy 2.9 888b72288bca 2 weeks ago 538.1MB 73.86MB 464.3MB 1
jdy_to_db 1.9 c345c4e15c1a 7 months ago 587.7MB 73.86MB 513.9MB 0
bot_etl  2.3 020d41691ec7 7 months ago 508.6MB 73.86MB 434.8MB 1
ubuntu  20.04 adafef2e596e 8 months ago 73.86MB 73.86MB 0B  0
Containers space usage:
CONTAINER ID IMAGE  COMMAND   LOCAL VOLUMES SIZE CREATED STATUS NAMES
011a83deceac jdy_extension:1.7 "supervisord -n -c /…" 1  17.8kB 15 hours ago Up 15 hours jdy_extension_1.7
712f65baea85 kudu_to_jdy:2.9 "supervisord -n -c /…" 1  36.9kB 2 weeks ago Up 45 hours kudu_to_jdy_2.9
a5fa673086c8 bot_etl:2.3  "supervisord -n -c /…" 1  72.8kB 7 months ago Up 45 hours bot_etl_2.3
Local Volumes space usage:
VOLUME NAME  LINKS SIZE
jdy_extensions_logs 1  658.5kB
kudu_to_jdy_kudu_to_jdy 1  128kB
bot_etl_bot_etl  1  724kB
Build cache usage: 0B
CACHE ID CACHE TYPE SIZE CREATED LAST USED USAGE SHARED
# 刪除所有沒有在用的volume
>>> docker volume rm $(docker volume ls -q)

補充:記一次docker磁盤占用100%情況

在執行git pull origin master的時候,拋出如下錯誤:

error: RPC failed; HTTP 500 curl 22 The requested URL returned error: 500

fatal: the remote end hung up unexpectedly

百度之後好多說是nginx文件大小限制,但是查看配置文件,大小限制已經設置到瞭500M,應該不是該原因造成的:

server
{
  listen 80;
  server_name localhost;
  client_max_body_size 500M;
  location / {
    proxy_redirect off;
    proxy_set_header Host $host;
    proxy_set_header X-Real-IP $remote_addr;
    proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for;
    proxy_pass http://localhost:180;
  }
}

然後使用df -h查看磁盤情況,發現data2磁盤已經100%(docker存儲在這裡):

使用docker system prune -a釋放瞭很多空間,問題解決。

註意:使用上述命令會刪除一下內容

WARNING! This will remove:
    - all stopped containers
    - all networks not used by at least one container
    - all images without at least one container associated to them
    - all build cache
Are you sure you want to continue? [y/N] y

以上為個人經驗,希望能給大傢一個參考,也希望大傢多多支持WalkonNet。如有錯誤或未考慮完全的地方,望不吝賜教。

推薦閱讀: