Proxmox 5.1 : zfs sync error

[dea]

Hi all !!

On a 3 nodes cluster (locale storage in ZFS) on recent PX 5.1 I have this error:

*********
2017-10-24 16:04:00 100-0: start replication job
2017-10-24 16:04:03 100-0: guest => VM 100, running => 4587
2017-10-24 16:04:03 100-0: volumes => zfs-pool:vm-100-disk-1
2017-10-24 16:04:05 100-0: create snapshot '__replicate_100-0_1508853840__' on zfs-pool:vm-100-disk-1
2017-10-24 16:04:05 100-0: incremental sync 'zfs-pool:vm-100-disk-1' (__replicate_100-0_1508502600__ => __replicate_100-0_1508853840__)
2017-10-24 16:04:05 100-0: internal error: Invalid argument
2017-10-24 16:04:05 100-0: command 'zfs send -Rpv -I __replicate_100-0_1508502600__ -- zfs-pool/vm-100-disk-1@__replicate_100-0_1508853840__' failed: got signal 6
2017-10-24 16:04:06 100-0: zfs-pool/vm-100-disk-1@__replicate_100-0_1508502600__ name zfs-pool/vm-100-disk-1@__replicate_100-0_1508502600__ -
2017-10-24 16:04:06 100-0: cannot receive: failed to read from stream
2017-10-24 16:04:06 100-0: command 'zfs recv -F -- zfs-pool/vm-100-disk-1' failed: exit code 1
2017-10-24 16:04:06 100-0: delete previous replication snapshot '__replicate_100-0_1508853840__' on zfs-pool:vm-100-disk-1
2017-10-24 16:04:06 100-0: end replication job with error: command 'set -o pipefail && pvesm export zfs-pool:vm-100-disk-1 zfs - -with-snapshots 1 -snapshot __replicate_100-0_1508853840__ -base __replicate_100-0_1508502600__ | /usr/bin/ssh -o 'BatchMode=yes' -o 'HostKeyAlias=proxmox1' root@192.168.1.217 -- pvesm import zfs-pool:vm-100-disk-1 zfs - -with-snapshots 1 -base __replicate_100-0_1508502600__' failed: exit code 255
*********

 

[dendi]

Hi dea,
Did you upgrade from 5.0 or it's a new install?
I have a cluster with ZFS replica and I'm planning to upgrade to 5.1...

 

[fabian]

could you please post (all of) the following:

• pveversion -v
• zfs list -t all -r zfs-pool
• zpool history zfs-pool
  
• pvesr list
• pvesr status (on all nodes)
• qm config 100
• journalctl --since "2017-10-24 16:03:00" --until "2017-10-24 16:08:00"

 

[dea]

OK, I've solved the problem.
Node 1 and 2 are update and run with new kernel, node 3 update but run on old (5.0) kernel.
I've seen zfs version is changed.
After reboot node 3, all function fine !

Thanks !!!!