have 2 test system running last Nexenta 4.0.3-FP4community .Master and slave ,

mbuffer installed on both,ssh trust is ok between hosts

when running znapzend on master replication snapshot did not ship to slave system,here my znapzend znapzend debug log:

could you please somebody explain where is problem?

./znapzend --debug --runonce=oztssdpool/VMS
[Mon Jun  1 10:39:22 2015] [info] refreshing backup plans...
[Mon Jun  1 10:39:23 2015] [info] found a valid backup plan for oztssdpool/VMS...
[Mon Jun  1 10:39:23 2015] [debug] snapshot worker for oztssdpool/VMS spawned (28160)
[Mon Jun  1 10:39:23 2015] [info] creating recursive snapshot on oztssdpool/VMS
# zfs snapshot -r oztssdpool/VMS@2015-06-01-103923
[Mon Jun  1 10:39:23 2015] [debug] snapshot worker for oztssdpool/VMS done (28160)
[Mon Jun  1 10:39:23 2015] [debug] send/receive worker for oztssdpool/VMS spawned (28163)
[Mon Jun  1 10:39:23 2015] [info] starting work on backupSet oztssdpool/VMS
# zfs list -H -r -o name oztssdpool/VMS
[Mon Jun  1 10:39:23 2015] [debug] sending snapshots from oztssdpool/VMS to root@ozteksas1:oztsas1pool/VMS
# zfs list -H -o name -t snapshot -s creation -d 1 oztssdpool/VMS
# ssh -o Compression=yes -o CompressionLevel=1 -o Cipher=arcfour -o batchMode=yes -o ConnectTimeout=30 root@ozteksas1 zfs list -H -o name -t snapshot -s creation -d 1 oztsas1pool/VMS
# ssh -o Compression=yes -o CompressionLevel=1 -o Cipher=arcfour -o batchMode=yes -o ConnectTimeout=30 root@ozteksas1 zfs list -H -o name -t snapshot -s creation -d 1 oztsas1pool/VMS
[Mon Jun  1 10:39:24 2015] [debug] receive process on ozteksas1 spawned (28169)
# zfs send oztssdpool/VMS@2015-06-01-103923|/opt/csw/bin/mbuffer -q -s 128k -W 60 -m 256M -O 'ozteksas1:31338'
# ssh -o Compression=yes -o CompressionLevel=1 -o Cipher=arcfour -o batchMode=yes -o ConnectTimeout=30 root@ozteksas1 '/opt/csw/bin/mbuffer -q -s 128k -W 60 -m 256M -4 -I 31338|zfs recv -F oztsas1pool/VMS'
***cannot receive new filesystem stream: dataset is busy
[Mon Jun  1 10:39:28 2015] [debug] receive process on ozteksas1 done (28169)
Mojo::Reactor::Poll: Read failed: Event "close" failed:  at /opt/znapzend/bin/../lib/ZnapZend/ZFS.pm line 380.***
# ssh -o Compression=yes -o CompressionLevel=1 -o Cipher=arcfour -o batchMode=yes -o ConnectTimeout=30 root@ozteksas1 zfs list -H -o name -t snapshot -s creation -d 1 oztsas1pool/VMS
[Mon Jun  1 10:39:29 2015] [debug] cleaning up snapshots on root@ozteksas1:oztsas1pool/VMS
# zfs list -H -o name -t snapshot -s creation -d 1 oztssdpool/VMS
[Mon Jun  1 10:39:29 2015] [debug] cleaning up snapshots on oztssdpool/VMS
[Mon Jun  1 10:39:29 2015] [info] done with backupset oztssdpool/VMS in 6 seconds
[Mon Jun  1 10:39:29 2015] [debug] send/receive worker for oztssdpool/VMS done (28163)
  • that dataset error seems odd ... does send/recv with the command line above work when you execute it by hand ? – Tobi Oetiker Jun 1 '15 at 11:51
  • master system zfs send oztssdpool/FILES@2015-06-01-140928|/opt/csw/bin/mbuffer -q -s 128k -W 60 -m 256M -O 'ozteksas1:31338' slave system /opt/csw/bin/mbuffer -q -s 128k -W 60 -m 256M -4 -I 31338|zfs recv -F oztsas1pool/FILES cannot receive new filesystem stream: dataset is busy did you mean like this? – expertaz Jun 1 '15 at 12:35
  • yes Tobi same result when I tried by hand ,but it works with ssh by hand when I remove destination zvol from slave system – expertaz Jun 1 '15 at 14:22

I found answer myself, when destitaion zvol shared(lun mapping) this error occurs,

after unsharing destination zvol,znapzend working since yesterday.

  • Same here, I was exporting the zvol with iSCSI and even though it wasn't being used during the send/recv. In my case (using FreeNAS) I temporarily removed the iSCSI Extent and it worked like a charm. – Chris Tonkinson Aug 28 '15 at 13:23

Your Answer

By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Not the answer you're looking for? Browse other questions tagged or ask your own question.