UI Bug in Datacenter > Options > Migration and Replication Settings

christian.g

Well-Known Member
Jun 4, 2020
54
33
58
Hello,

the dropdown to select a network for Migration/Replication doesn't show/select the networks but the IP for the node one is logged in.

Example on node 1.
All networks are X.X.X.0/24
And configured IPs are X.X.X.1/24

1757708785147.png

Selecting one of those entries sets the IP (ex. 10.0.110.1/24) instead of the network (10.0.110.0/24) which results in a wrong configuration, if users don't pay attention an manually correct the select entry.
 
Yeah, those should be normalized to the network, thanks for your report. But FWIW, the backend still does the right thing, so it's mostly confusing UX.

Might be good to create a report in our Bugzilla issue tracker to better ensure we do not lose track of this. https://bugzilla.proxmox.com/
 
Thanks for the info.

Right now i'm having a hard time fighting a few heavy problems with PVE 9.
  • All VMs on node1 lost network connectivity and i had to reboot the node. After it came back and VMs started migrating back the next problems showed up on a another node (see below)
  • Two different migration problems on node2 while migrating back to node1
    • Cleanup after stopping VM failed - org.freedesktop.DBus.Error.NoReply: Did not receive a reply. Possible causes include: the remote application did not send a reply, the message bus security policy blocked the reply, the reply timeout expired, or the network connection was broken.
      • Not sure if it's harmful cause the VMs got relocated in the end.
    • ERROR: migration aborted (duration 00:00:00): VM 141 qmp command 'block-dirty-bitmap-add' failed - Bitmap already exists: repl_virtio
      • This was a show stopper as the node wasn't able to cleanup the existing bitmap and thus couldn't relocate the VM (endless loop), nor was i able to reboot node2 due to this. Had to kill the qemu process for that VM to be able to proceed. This VM is a database server with replicated disks. Probably lost data due to this.
I know i should open a new thread/bugreport but it's late now an i need to watch the infrastructure for more problems.