On Tue, Dec 20, 2016 at 4:18 PM, Devin Acosta <devin@pabstatencio.com> wrote:

One interesting thing I noticed today after looking at the logs from the 'ovn-controller' on the oVIRT nodes, is node1 is logging like crazy, the file is over 1.2GB in size already, seems to be looping like crazy? What should i check to see why it is logging and maxing CPU like this?

2016-12-20T21:16:17.517Z|91605|binding|INFO|Changing chassis for lport d2e19c41-f8ff-49d2-bd8e-01d833e59a59 from 252778b5-bc63-486d-8d82-c3e4350ee650 to c0dc0909-4274-400d-826e-82348fee47e9.
2016-12-20T21:16:17.521Z|91606|poll_loop|INFO|Dropped 5666 log messages in last 6 seconds (most recently, 0 seconds ago) due to excessive rate
2016-12-20T21:16:17.521Z|91607|poll_loop|INFO|wakeup due to [POLLIN] on fd 12 (172.20.192.73:54710<->172.20.192.77:6642) at lib/stream-fd.c:155 (94% CPU usage)
2016-12-20T21:16:17.521Z|91608|binding|INFO|Changing chassis for lport d2e19c41-f8ff-49d2-bd8e-01d833e59a59 from 252778b5-bc63-486d-8d82-c3e4350ee650 to c0dc0909-4274-400d-826e-82348fee47e9.

Check "ovn-sbctl show" to figure out which host is chassis 252778b5-bc63-486d-8d82-c3e4350ee650. See if the other host has similar entries in its log.

This could happen if a port is configured on two hosts at the same time. They will constantly fight over which chassis owns it.

Russell Bryant