You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The "smart" send is likely to send a zfs-auto-snap_frequent one. Later once the zfs-auto-snap_frequent get deleted from local host, zfsbackup will get confused and refuse to backup any further.
The text was updated successfully, but these errors were encountered:
It's somewhat out of scope for this project to handle all cases of how you manage your snapshots, you can script execution of this project manually according to how you manage snapshots.
How would you like to handle these cases? What's the expected behavior?
One: zfsbackup-go has a --snapshotPrefix param so you can tell it to e.g. only consider the daily or hourly snapshots. I submitted this feature precisely to deal with this scenario.
Two: It is probably better to have a separate set of snapshots that are only rotated when the backup runs. My strategy is to use zfsbackup-go as the snapshot label, and my wrapper script that runs the backup does this:
# create a new set of snapshots
zfs-auto-snapshot --label=zfsbackup-go //
zfs list -H -o name,com.sun:auto-snapshot | while read line; do
# run zfsbackup-go, unless com.sun:auto-snapshot is "false" ...
...
done
# delete old snapshots.
zfs-auto-snapshot --destroy-only --keep=1 --label=zfsbackup-go //
Side note: there is a bug in zfs-auto-snapshot where you need to set --keep=2 unless you have patched around it. See zfsonlinux/zfs-auto-snapshot#121
The "smart" send is likely to send a
zfs-auto-snap_frequent
one. Later once thezfs-auto-snap_frequent
get deleted from local host, zfsbackup will get confused and refuse to backup any further.The text was updated successfully, but these errors were encountered: