exoframe-server
exoframe-server copied to clipboard
docker image prune -a
I would like to reclaim more space. In my case it was like 4GB after couple month of daily deploys More info https://github.com/apocas/dockerode/issues/429
docker system df
TYPE TOTAL ACTIVE SIZE RECLAIMABLE
Images 18 6 5.501GB 4.639GB (84%)
Containers 6 6 64.72MB 0B (0%)
Local Volumes 3 3 50.92MB 0B (0%)
Build Cache 0 0 0B 0B
And after pruning with flag -a
TYPE TOTAL ACTIVE SIZE RECLAIMABLE
Images 6 6 1.587GB 0B (0%)
Containers 6 6 64.72MB 0B (0%)
Local Volumes 3 3 50.92MB 0B (0%)
Build Cache 0 0 0B 0B
Coverage remained the same at 87.462% when pulling 6f85cacbd1a311450ba71e2e78529a865b7b49e7 on FDiskas:patch-1 into 5725c91c4ba759011cd17c6d11a3dc420c632f89 on exoframejs:master.
This changes the behavior quite a bit. It should be behind a flag IMO 🤔
This is what I'm thinking:
exoframe is not going to mimic/forward docker commands with all his additional params.
the same as exoframe rm container - impossible to remove multiple containers at once - just because it's not going to replace docker functionality - but rather then a helper to do some common stuff without actually connecting to the server. Lets allow exoframe do its job (cleanup as much as possible) - on dedicated server size is matter. But in other cases like if end user is very delicate - he can cleanup connected to the server and using docker with all params or without params as he wants.
I'm not saying exoframe should mimic all docker commands, I'm just saying this particular one should be behind an opt-in flag. Removing all dangling images has quite a severe impact (on docker builds, starting services, etc) - so IMO it shouldn't be default, it needs a flag.
This flag should look the same -a?
@FDiskas yeah, --all (short -a) would be perfect