[pve-devel] replica problems

Dominik Csapak d.csapak at proxmox.com
Thu Jun 1 11:01:30 CEST 2017


hi,

i tested the replica patches a bit and noticed following problems:

1. when we have a job with a failcount > 0, we can't delete/disable the 
job (only manually deleting from /etc/pve/replication.cfg works)

2. the error message saved is not really helpful, an example:

command 'set -o pipefail && pvesm export local-zfs:vm-100-disk-1 zfs - 
-snapshot replicate_replica01_1496307002_snap -with-snapshots 1 | 
/usr/bin/ssh -o 'BatchMode=yes' -o 'HostKeyAlias=pve-local-03' 
root at 192.168.20.63 -- pvesm import local-zfs:vm-100-disk-1 zfs - 
-with-snapshots 1' failed: exit code 255

this does not really tell me, why it does not work, and manually 
executing this does also not work, because the snapshot does not exist 
anymore

3. we do not have a cron entry for pvesr run (yet? by design?)

4. following scenario does not work (and i don't know why)

3 nodes (node01-node03)

vm 100 on node01, replica job to node03

migrate vm 100 to node02 (works without problems)

replica job stops working

expectation: vm100 gets now replicated from node02 -> node03




More information about the pve-devel mailing list