I am working on a complimentary script that does log-shipping style replication to a backup volume for highly efficient incremental backups. This relies on "zfs send -R" and that the sets of snapshots existing on the source and backup destination overlap by at least one snapshot. I plan to use snapshot clones to place a "checkpoint" reservation on a snapshot that will be required for future backups.
In order to keep zfs-auto-snapshot from destroying a snapshot necessary for a future incremental backup replication stream, I will create an unmounted clone of that snapshot, and destroy any prior clones when the backup completes.
Any snapshots destroyed by the zfs-auto-snapshot "zfs destroy -d" will be destroyed as the dependent clones are destroyed. Without the -d option, zfs-auto-snapshot would fail to destroy the old snapshot with a checkpoint clone, and it would persist until later snapshot jobs destroy it in subsequent runs. Since monthlies and weeklies are relatively infrequent, deferred destruction will reap the snapshots opportunistically as soon as the "checkpoint" clones are gone. Intermediate snapshots between the "checkpoint" reserved snapshots and the recent snapshots preserved by the zfs-auto-snapshot.sh will still be immediately destroyed.
Resolve this lintian error:
E: zfs-auto-snapshot: bad-permissions-for-etc-cron.d-script etc/cron.d/zfs-auto-snapshot 0755 != 0644
W: zfs-auto-snapshot: executable-not-elf-or-script etc/cron.d/zfs-auto-snapshot
This should also resolve a related failure on Arch Linux.
Closes: zfsonlinux/zfs-auto-snapshot#4
Thanks: @bitloggerig
Thanks: @scottj97
Implement a `--fast` switch that uses the optimized `zfs list`
invocation instead of the slower regular form.
If the optimized form is generally better, then use it as the
default and reverse the switch to `--slow` sometime later.
Issue: zfsonlinux/zfs-auto-snapshot#16
zfs list is very slow if it needs to retrieve other properties than name. See zfsonlinux/zfs#450 for reference.
This change uses only the name of the snapshot and uses awk to extract the snapshot date from the name and sort to sort it:
1. get all snapshot names, sorted by name
2. use grep to filter out any snapshot, that isn't correctly prefixed (maybe a check for the correct date format could also be added here)
3. use awk to extract the date and put the date before the snapshot
4. use sort to reverse sort this, first by date, then by snapshot name
5. use awk to remove the date, so that just the snapshot name is left
This significally speeds it up on my system (/root/zfs-auto-snapshot is the changed one, running with -n to get only the times for snapshot retrieval):
root@tatooine:/root# time /root/zfs-auto-snapshot -d -v // --label=frequent --keep=4 -n
Debug: Including rpool for regular snapshot.
Debug: Including rpool/DATA for regular snapshot.
Debug: Including rpool/DATA/ldap for recursive snapshot.
Debug: Including rpool/DATA/postgresql for recursive snapshot.
Debug: Including rpool/DATA/vm for regular snapshot.
Debug: Including rpool/DATA/vm/alderaan.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/bespin.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/coruscant.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/dagobah.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/dantooine.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/dev.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/monitor.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/office.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/test.example.net for recursive snapshot.
Debug: Including rpool/ROOT for recursive snapshot.
Debug: Excluding rpool/ROOT/ubuntu-1 because rpool/ROOT includes it recursively.
Doing regular snapshots of rpool rpool/DATA rpool/DATA/vm
Doing recursive snapshots of rpool/DATA/ldap rpool/DATA/postgresql rpool/DATA/vm/alderaan.example.net rpool/DATA/vm/bespin.example.net rpool/DATA/vm/coruscant.example.net rpool/DATA/vm/dagobah.example.net rpool/DATA/vm/dantooine.example.net rpool/DATA/vm/dev.example.net rpool/DATA/vm/monitor.example.net rpool/DATA/vm/office.example.net rpool/DATA/vm/test.example.net rpool/ROOT
Doing a dry run. Not running these commands...
zfs snapshot -o com.sun:auto-snapshot-desc='-' 'rpool@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy 'rpool@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' 'rpool/DATA@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy 'rpool/DATA@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' 'rpool/DATA/vm@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy 'rpool/DATA/vm@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/ldap@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/ldap@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/postgresql@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/postgresql@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/alderaan.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/alderaan.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/bespin.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/bespin.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/coruscant.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/coruscant.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/dagobah.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/dagobah.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/dantooine.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/dantooine.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/dev.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/dev.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/monitor.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/monitor.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/office.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/office.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/test.example.net@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/DATA/vm/test.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/ROOT@zfs-auto-snap_frequent-2013-04-16-2344'
zfs destroy -r 'rpool/ROOT@zfs-auto-snap_frequent-2013-04-16-2315'
@zfs-auto-snap_frequent-2013-04-16-2344, 15 created, 15 destroyed, 0 warnings.
real 0m6.936s
user 0m0.076s
sys 0m0.184s
root@tatooine:/root# time /sbin/zfs-auto-snapshot -d -v // --label=frequent --keep=4 -n
Debug: Including rpool for regular snapshot.
Debug: Including rpool/DATA for regular snapshot.
Debug: Including rpool/DATA/ldap for recursive snapshot.
Debug: Including rpool/DATA/postgresql for recursive snapshot.
Debug: Including rpool/DATA/vm for regular snapshot.
Debug: Including rpool/DATA/vm/alderaan.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/bespin.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/coruscant.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/dagobah.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/dantooine.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/dev.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/monitor.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/office.example.net for recursive snapshot.
Debug: Including rpool/DATA/vm/test.example.net for recursive snapshot.
Debug: Including rpool/ROOT for recursive snapshot.
Debug: Excluding rpool/ROOT/ubuntu-1 because rpool/ROOT includes it recursively.
Doing regular snapshots of rpool rpool/DATA rpool/DATA/vm
Doing recursive snapshots of rpool/DATA/ldap rpool/DATA/postgresql rpool/DATA/vm/alderaan.example.net rpool/DATA/vm/bespin.example.net rpool/DATA/vm/coruscant.example.net rpool/DATA/vm/dagobah.example.net rpool/DATA/vm/dantooine.example.net rpool/DATA/vm/dev.example.net rpool/DATA/vm/monitor.example.net rpool/DATA/vm/office.example.net rpool/DATA/vm/test.example.net rpool/ROOT
Doing a dry run. Not running these commands...
zfs snapshot -o com.sun:auto-snapshot-desc='-' 'rpool@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy 'rpool@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' 'rpool/DATA@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy 'rpool/DATA@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' 'rpool/DATA/vm@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy 'rpool/DATA/vm@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/ldap@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/ldap@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/postgresql@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/postgresql@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/alderaan.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/alderaan.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/bespin.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/bespin.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/coruscant.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/coruscant.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/dagobah.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/dagobah.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/dantooine.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/dantooine.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/dev.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/dev.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/monitor.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/monitor.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/office.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/office.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/DATA/vm/test.example.net@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/DATA/vm/test.example.net@zfs-auto-snap_frequent-2013-04-16-2315'
zfs snapshot -o com.sun:auto-snapshot-desc='-' -r 'rpool/ROOT@zfs-auto-snap_frequent-2013-04-16-2348'
zfs destroy -r 'rpool/ROOT@zfs-auto-snap_frequent-2013-04-16-2315'
@zfs-auto-snap_frequent-2013-04-16-2348, 15 created, 15 destroyed, 0 warnings.
real 3m30.995s
user 0m0.152s
sys 0m0.792s
root@tatooine:/root#
I'm not an awk god, so I tried a bit until I got a working version. There might be better ways. I also don't know if this catches every adge case, but it is a start for improvement. :)
Use $ZPOOL_STATUS instead of $ZFS_STATUS for pool exclusions.
A mistaken variable name made the pool exclusion logic a no-op and
broke the --skip-scrub switch.
Also correct a trivial comment typo.
Split the cron file so that the hourly, daily, weekly, and monthly
zfs automatic snapshots still happen even if the system is offline
temporarily during the scheduled event, such as during a reboot.
On Debian systems, if anacron is installed, then it is used to run
the /etc/cron.{hourly,daily,weekly,monthly} directories, but not
the /etc/cron.d directory. This means that /etc/cron.d jobs are not
run if the system is offline when crond would usually invoke them.
Set the com.sun:auto-snap-desc property on each snapshot to an
arbitrary value. On Solaris, this property is set to the dash
character by default -- which is convention for NULL -- but it
can contain an FMRI status comment or other user data.
