Hi! I have LVM thin pool on my server(srv1) which use for lxc containers.
When i create a backup containers i use
lxc snapshot name_container
lxc publish name_container/name_snapshot
lxc image export name_image to the backup disks (large storage)
a few days ago my server is broken, but i have a backups
i restore my backups on another server(srv2) with DIR storage instead lvm and all OK.
yesterday i repair my first server(srv1) and setup lvm and lxd
I create lxc lvm pool, pushing containers from srv2 to srv1 and started it - all ok
when i start backup script, my root partition (/) on srv1 growing when containers publishing
i have on root partirion only 50gb free space and and backup script failed, because image of container bigger than free space on root /
BUT, when srv1 has been not broken (week ago) i see my root partition is clear when i backup containers every night. I don’t know where is stores images, but i exactly know my root partition is not growing at all
Why publishing image go to the /var/lib/lxd/images instead /var/lib/lxd/storage_pools/images?
Can i configure publishing images to the LVM storage pool instead default location?
How to change local LXD image store - to storage pool image store
I asked because it’s works on my srv1 (week ago)
reason of broking my srv1 is metedata on lvm pool 100% maybe images are published in metedata all time? and it’s overspaced? I don’t know