Thursday, 2021-11-18

*** rlandy|ruck|bbl is now known as rlandy|ruck02:40
*** rlandy|ruck is now known as rlandy|out02:44
*** queria is now known as Guest615803:25
*** queria is now known as Guest616003:30
*** soniya29|ruck is now known as soniya29|ruck|lunch07:13
*** soniya29|ruck|lunch is now known as soniya29|ruck08:11
jpichi all, if I have 2 availability zones with 3 controlers on each, do I need a 6 node mariadb cluster or is it better to have 2 clusters with 3 nodes each? thanks10:04
fricklerjpic: if you want each zone to be able to work when the other one is completely down, then you need 2 separate clusters. with a 6 node cluster, you'd still need 4 nodes to be online in order to have quorum10:08
jpicthank you very much frickler 10:09
jpicfrickler: if I have 6 nodes + 1 galera arbitrator, and one zone is completely down, am I fine too?10:19
*** rlandy|out is now known as rlandy|ruck11:11
*** marlinc is now known as Guest621517:47
*** tkajinam is now known as Guest622018:30
SatheeshGreetings All - We are getting an issue when we have Control Plane in Ussuri version (21.2.10) and Compute plane is in Train (20.2.1) version, we are getting Virtual Interface Creation Error while provisioning the VM? nova.exception.VirtualInterfaceCreateException: Virtual Interface creation failed18:43
SatheeshDid anyone faced such issue? ERROR neutron.plugins.ml2.drivers.agent._common_agent [req-3806c12d-6522-4b6a-b575-33d2faf0b859 - - - - -] Error in agent loop. Devices info: {xxx}: ValueError: too many values to unpack (expected 2)18:44
fricklerSatheesh: can you show the full traceback from the error message on also, those are nova version numbers, which versions of neutron do you have in place?19:01
SatheeshThanks Frickler - nova-compute version is 20.4.2 and neutron-linuxbridge-agent version is 15.3.1.dev2619:23
SatheeshComplete Error log snippet is attached here..
fricklerSatheesh: neutron did mess up things there a bit, you need to make sure that you have exactly the same version on the controller or upgrade to a later version in train
fricklerbest ask in #openstack-neutron if you need further help with that, although also train is essentially EOL by now19:48
SatheeshThanks Frickler... I noticed that neutron bug earlier.. will reach out to #openstack-neutron channel to know the actual fix for this.. As part of our upgrade plan, we used to upgrade Control plane separately and then after few days Compute plane will be upgraded.. During that duration, we need to ensure the provisioning should work properly without any issue.. We didn't find similar issue when we were in Train control plane and Stein compute plane20:41

Generated by 2.17.2 by Marius Gedminas - find it at!