Open mariusrak opened 4 years ago
Could you please show me the output from running these commands on the server: docker system df --verbose
and docker ps -qa | xargs docker inspect --format='{{.LogPath}}' | xargs ls -hl
.
Sorry, meanwhile I cleared and erased everything. But as soon as it occurs again I'll post the outputs. For now, I'm just happy, that it isn't normal. Thanks.
I've deployed an app to AWS cca half a year ago. Then after a half year it was not accessible from internet, so I've connected to ssh to see what is going on and found out that docker ate whole 8gb disc, leaving 0 bytes free. This caused docker to crash and unable to start.
The application was doing basically nothing, there was no traffic since it was only an experiment. Yet it still ate whole disc. Most of data is stored in
/var/lib/docker/overlay2/
. After some googling I've tried running recommendeddocker system prune --all
. This helped a liittle but theoverlay2
directories are still too large.I have no idea where this data came from or how to debug it, but I think such stuff, where something just grows should not happen, right?
mup version: 1.4.5 meteor version 1.9-beta.0 deployed to aws t2.micro Ubuntu 18.04.2 LTS