numerous-motorcycle-43166
04/26/2023, 2:56 PMskaffold
and it works most of the time, however after 2-3 reloads from my code changes, the node becomes unusable due to disk pressure. I'm using linux with rancher desktop 1.8.1. Has anyone ran into this before or have any suggestions to try to help avoid this disk pressure issue?fast-garage-66093
04/27/2023, 3:32 AM/opt/rancher-desktop/resources/resources/linux/lima/bin/qemu-img resize ~/.local/share/rancher-desktop/lima/0/diffdisk +10G
skaffold
has to delete images no longer needed, but it sounds like you should frequently prune the image store.kubelet
is going to do some garbage collection when a threshold is reached, but maybe that threshold is too low, so you exhaust the disk before GC can kick in.numerous-motorcycle-43166
04/27/2023, 12:59 PMdocker system prune
it does show 90+ GB. I've been running skaffold with --no-prune=false --cache-artifacts=false
which should prune the images more often; so hopefully that helps toofast-garage-66093
04/27/2023, 4:25 PMnumerous-motorcycle-43166
04/27/2023, 5:37 PMqemu-img
in that path though, only `limactl`:
❯ ll /opt/rancher-desktop/resources/resources/linux/lima/bin/
total 19844
drwxr-xr-x 2 root root 4096 Apr 19 12:57 ./
drwxr-xr-x 4 root root 4096 Apr 19 12:57 ../
-rwxr-xr-x 1 root root 20312064 Feb 28 16:52 limactl*
fast-garage-66093
04/27/2023, 5:39 PMqemu-img
on the PATH
numerous-motorcycle-43166
04/27/2023, 5:40 PMqemu-img resize ~/.local/share/rancher-desktop/lima/0/diffdisk +10G
worked. Will test the results of a bigger size now.fast-garage-66093
04/27/2023, 5:42 PMnumerous-motorcycle-43166
04/27/2023, 5:49 PMfast-garage-66093
04/27/2023, 5:54 PMnumerous-motorcycle-43166
05/02/2023, 3:45 PM--no-prune=false --cache-artifacts=false
options with skaffold has been pretty reliable