Hi, yesterday I had problems “downgrading” to official Debian lxd package and I solved with @tomp help (MANY Thanks!).
Now Debian package is installed correctly and daemon starts without errors.
I need to lxd init though (as stated by Debian installation).
I tried reusing my old setup by giving:
cat old_config.yaml | lxd init --preseed
but it rises error:
Error: Failed to create storage pool "default": Provided ZFS pool (or dataset) isn't empty, run "sudo zfs list homez/lxd" to see existing entries
using said command shows an empty filesystem occupying a lot of space, OTOH plain sudo zfs list shows sub-filesystems containing a lot of stuff, including all my containers:
mcon@cinderella:~$ sudo zfs list
NAME USED AVAIL REFER MOUNTPOINT
homez 2.55T 2.61T 406G /homez
homez/etherpad 248M 2.61T 248M /homez/etherpad
homez/lxd 47.3G 2.61T 140K none
homez/lxd/containers 41.5G 2.61T 140K none
homez/lxd/containers/Spoke 6.93G 2.61T 7.31G legacy
homez/lxd/containers/lucid 1.64G 2.61T 2.18G legacy
homez/lxd/containers/lxdMosaic 1.71G 2.61T 2.09G legacy
homez/lxd/containers/nrf-builder 4.31G 2.61T 4.71G legacy
homez/lxd/containers/nrf-devel 8.88G 2.61T 9.42G legacy
homez/lxd/containers/redmine 818M 2.61T 1.18G legacy
homez/lxd/containers/seeme-builder 4.41G 2.61T 4.92G legacy
homez/lxd/containers/synology-builder 11.3G 2.61T 11.7G legacy
homez/lxd/containers/yocto-builder 1.59G 2.61T 2.12G legacy
homez/lxd/custom 140K 2.61T 140K none
homez/lxd/deleted 4.40G 2.61T 140K none
homez/lxd/deleted/containers 140K 2.61T 140K none
homez/lxd/deleted/custom 140K 2.61T 140K none
homez/lxd/deleted/images 4.40G 2.61T 140K none
homez/lxd/deleted/images/013beea918448a9fc4ae5419b132c718cac3ac58f86a024ac7506d9cdbd2e0fd 394M 2.61T 394M legacy
homez/lxd/deleted/images/0db04e4645814398aca82cf3d99098dc3d317c60c9d4258c9808bd91fd67454a 570M 2.61T 570M none
homez/lxd/deleted/images/4de6a2df17edbd2cd8a9b1dc619b4a5ac71f97755a747ca401b1c19c1870b04e 556M 2.61T 556M none
homez/lxd/deleted/images/59b17fef4047dad12df4a71ed00e348816eec8260701fd78b90b57bede348267 562M 2.61T 562M none
homez/lxd/deleted/images/78669956eba4767c3bd4ef99538d8eac6d179b14b0232d7c58abc4e7dbe48dc9 396M 2.61T 396M legacy
homez/lxd/deleted/images/7dbe2f94667748fb2355e5e46b75f20a510277ef96aad33be743e9b75ec1d301 594M 2.61T 594M legacy
homez/lxd/deleted/images/b89dd4e241ca07a7f8cdef6109aaa2a02458a19ec80baaa2e3e1974d74eb7852 396M 2.61T 396M legacy
homez/lxd/deleted/images/d3402556d2856568217a4f6536f9005567d02957982022a6e917d9f6da77359f 470M 2.61T 470M legacy
homez/lxd/deleted/images/f71fea6a1e44034abae9c63a0e625de321e144c54bf478d8aefa3035465205e5 566M 2.61T 566M none
homez/lxd/deleted/virtual-machines 140K 2.61T 140K none
homez/lxd/images 1.41G 2.61T 140K none
homez/lxd/images/9d10e8b9e175cdc5d8a331a439963f002c6855d35ce52b27d3cdb9c68d76c4c4 396M 2.61T 396M legacy
homez/lxd/images/d75e77c8452487d390c8ee86629f147fee110d83fb79b053bc9ad107fce7aa1c 574M 2.61T 574M none
homez/lxd/images/f4451070c9990ee862b9fe6bac4c06f62c7cdc387783ce8724dc6309989d0b42 469M 2.61T 469M legacy
homez/lxd/virtual-machines 140K 2.61T 140K none
homez/mauro 140K 2.61T 140K /home/mauro
homez/mcon 2.10T 2.61T 2.10T /home/mcon
mcon@cinderella:~$ sudo mount -o zfsutil -t zfs homez/lxd /mnt
mcon@cinderella:~$ ls -la /mnt
total 5
drwxr-xr-x 2 root root 2 Dec 25 2021 .
drwxr-xr-x 20 root root 4096 May 13 23:40 ..
mcon@cinderella:~$ sudo umount /mnt
I could easily change the pool name and restart from scratch, but is there a way to recover my old containers?
This isn’t really vital because they are mostly used for trials and compilations in specific environments (just in one case I’m unsure if I saved all configs, but I can redo), but I would like to understand and learn, if possible.
I understand I should have done a lxc export before killing my old installation, but I didn’t (my bad, of course); is there another way to recover?
Thanks in advance
Mauro