oetiker / znapzend

zfs backup with remote capabilities and mbuffer integration.
www.znapzend.org
GNU General Public License v3.0
608 stars 138 forks source link

Znapzend executes concurrent conflicting send/recvs #423

Closed Harvie closed 3 years ago

Harvie commented 5 years ago

In log i see lots of messages like this:

May 03 18:40:16 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-144500': Broken pipe
May 03 18:40:16 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-150000': Broken pipe
May 03 18:40:17 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-151500': Broken pipe
May 03 18:40:17 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-153000': Broken pipe
May 03 18:40:19 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-154500': Broken pipe
May 03 18:40:19 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-160000': Broken pipe
May 03 18:40:21 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-161500': Broken pipe
May 03 18:40:21 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-163000': Broken pipe
May 03 18:40:21 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-164500': Broken pipe
May 03 18:40:22 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-170000': Broken pipe
May 03 18:40:22 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-171500': Broken pipe
May 03 18:40:26 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-173000': Broken pipe
May 03 18:40:26 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-174500': Broken pipe
May 03 18:40:27 myhost znapzend[1983]: warning: cannot send 'tank/vps/subvol-106-disk-1@2019-04-22-180000': Broken pipe

I've found that znapzend-0.19.0 tries to replicate two different snapshots of the same dataset at the same time:

# pstree -a 1983
znapzend /usr/local/bin/znapzend --debug --autoCreation
  ├─znapzend /usr/local/bin/znapzend --debug --autoCreation
  │   └─sh -c...
  │       └─zfs send -I tank/vps/subvol-106-disk-1@2019-04-14-163000 tank/vps/subvol-106-disk-1@2019-05-03-183000
  └─znapzend /usr/local/bin/znapzend --debug --autoCreation
      └─sh -c...
          ├─ssh -o batchMode=yes -o ConnectTimeout=30 root@dsthost...
          └─zfs send -I tank/vps/subvol-106-disk-1@2019-04-11-061500 tank/vps/subvol-106-disk-1@2019-05-03-160000

I think this might have happened after enabling mbuffer and sending SIGHUP... Is this expected behavior? After restarting znapzend im back to single process...

oetiker commented 5 years ago

it look as if the ssh process has died ... in the first instance ... something seems wrong there

stale[bot] commented 3 years ago

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

Harvie commented 3 years ago

bump