[neutron] HA network issues on Rocky

Jean-Philippe Méthot jp.methot at planethoster.info
Wed Nov 18 19:03:47 UTC 2020


We’ve been running an Openstack setup in production for several years and we upgraded it to Rocky last May. We are now creating a router for a new project for the first time
since that upgrade and we’ve been running into issues with L3HA for that router. Essentially, the router for this project gets initialized with a faulty HA network. 

How do I know it’s faulty?

-Both Keepalived instances think they are master.
-Both HA network ports get the wrong MTU (9000 instead of 8958)
-Even if I manually set the proper MTU, VRRP packets don’t get received on the other node
-If I then disable and enable the router, MTU gets rolled back to 9000
-If I force the MTU for this port through the neutron database, it will also get rolled back to 9000 once the router gets recreated

This setup is using openvswitch and GRE tunnels. I have 7 other projects with L3HA routers running and they do not have this issue, even if I create new routers in them. They are also older projects created before the upgrade to Rocky. This problem is strictly isolated to this new project.

I’m not sure exactly what’s going on here and why Openstack isn’t taking encapsulation into account when automatically setting its HA network MTU for this. Any suggestion on how to figure out what’s happening behind the scene?

Jean-Philippe Méthot
Senior Openstack system administrator
Administrateur système Openstack sénior
PlanetHoster inc.
4414-4416 Louis B Mayer
Laval, QC, H7P 0G1, Canada
TEL : +1.514.802.1644 - Poste : 2644
FAX : +1.514.612.0678
CA/US : 1.855.774.4678
FR : 01 76 60 41 43
UK : 0808 189 0423

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-discuss/attachments/20201118/0f1c24e6/attachment-0001.html>

More information about the openstack-discuss mailing list