Weird Virsh/network issue/post upgrade issue

Community Support General
1

For the life of me I can’t figure this out - I’m about 24 hours in now lol

Mon Aug 30 16:13:03 2021 [Z0][VMM][I]: Command execution fail: cat << EOT | /var/tmp/one/vmm/kvm/attach_nic '01363b51-0bad-43b5-9017-347b56375485' '02:00:ac:10:64:02' 'minionebr' '-' 'bridge' 'one-88-1' 88 vhost2.maas
Mon Aug 30 16:13:03 2021 [Z0][VMM][E]: attach_nic: Command "virsh --connect qemu:///system attach-device 01363b51-0bad-43b5-9017-347b56375485 <(
Mon Aug 30 16:13:03 2021 [Z0][VMM][I]: cat <<EOT
Mon Aug 30 16:13:03 2021 [Z0][VMM][I]: <interface type='bridge'> <source bridge='minionebr'/> <mac address='02:00:ac:10:64:02'/> <target dev='one-88-1'/> </interface>
Mon Aug 30 16:13:03 2021 [Z0][VMM][I]: EOT
Mon Aug 30 16:13:03 2021 [Z0][VMM][I]: )" failed: error: Failed to attach device from /dev/fd/63
Mon Aug 30 16:13:03 2021 [Z0][VMM][I]: error: internal error: unable to execute QEMU command 'device_add': Bus 'pci.0' does not support hotplugging
Mon Aug 30 16:13:03 2021 [Z0][VMM][E]: Could not attach NIC 1 (02:00:ac:10:64:02) to 01363b51-0bad-43b5-9017-347b56375485
Mon Aug 30 16:13:03 2021 [Z0][VMM][I]: ExitCode: 1
Mon Aug 30 16:13:03 2021 [Z0][VMM][I]: Failed to execute virtualization driver operation: attach_nic.
Mon Aug 30 16:13:03 2021 [Z0][IPM][D]: Message received: ATTACHNIC FAILURE 88 Could not attach NIC 1 (02:00:ac:10:64:02) to 01363b51-0bad-43b5-9017-347b56375485

It doesn’t overly matter what I do in terms of the network setup itself, this seems to be an issue with the actual attachment operation

happy to provied any additional information - VM’s and hosts were working prior to updating to 6.0 - ‘error: internal error: unable to execute QEMU command ‘device_add’: Bus ‘pci.0’ does not support hotplugging’ appears to be the issue but my google fu isn’t directing me toward anything

2

Maybe it is not related, but did you run onehost sync to update the hosts scripts. In particular /var/tmp/one/vmm/kvm/attach_nic maybe outdated.

As the new versions uses rsync, it is safer to remove /var/tmp/one from the hypervisors and then perform a onehost sync --force…

Cheers

3

Hi,

I ended up throwing my hands up and rebuilding the entire cluster about an hour after my first post - apologies for not coming back and updating

Good part is that it was simply a case of reprovisioning the hosts in Maas and a quick run of an ansible playbook (after a couple of wee modifications) and my hosts were back up and running (I did lose VM’s but this is a lab env so that’s not a huge deal and these VM’s were ‘provisioned’ VM’s as well)

I do have another issue where almost everything is working but I now can’t migrate VM’s - but I think I need to set up some new datastores for that one

4

OK, Thanks for the update :slight_smile:

Good Luck!