Details
-
Bug
-
Resolution: Done
-
Major
-
None
-
6.5 SP1, 7.0, 7.3, 7.4, 7.5, 7.6
-
None
Description
Across all our hosts (20~) there are many Openvswitch processes running on the dom0s that regularly use a lot of CPU resources (between 30-100% of one Xeon E5-2660 v2 core).
11 3407 root -10 1270M 156M 8864 S 72.5 2.0 24h58:09 ovs-vswitchd unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:info -vfile:emer --mlockall --no-chdir --log-file=/var/log/ovs-vswitchd.log --pidfile=/var/run/openvswitch/ovs-vswitchd.pid --detach --monitor 5 4100 root -10 1270M 156M 8864 S 14.1 2.0 4h48:21 ovs-vswitchd unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:info -vfile:emer --mlockall --no-chdir --log-file=/var/log/ovs-vswitchd.log --pidfile=/var/run/openvswitch/ovs-vswitchd.pid --detach --monitor 3 4097 root -10 1270M 156M 8864 S 14.1 2.0 4h47:29 ovs-vswitchd unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:info -vfile:emer --mlockall --no-chdir --log-file=/var/log/ovs-vswitchd.log --pidfile=/var/run/openvswitch/ovs-vswitchd.pid --detach --monitor 9 4096 root -10 1270M 156M 8864 S 14.1 2.0 4h50:41 ovs-vswitchd unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:info -vfile:emer --mlockall --no-chdir --log-file=/var/log/ovs-vswitchd.log --pidfile=/var/run/openvswitch/ovs-vswitchd.pid --detach --monitor 13 4101 root -10 1270M 156M 8864 S 14.1 2.0 4h48:11 ovs-vswitchd unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:info -vfile:emer --mlockall --no-chdir --log-file=/var/log/ovs-vswitchd.log --pidfile=/var/run/openvswitch/ovs-vswitchd.pid --detach --monitor 6 4099 root -10 1270M 156M 8864 S 14.1 2.0 4h47:30 ovs-vswitchd unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:info -vfile:emer --mlockall --no-chdir --log-file=/var/log/ovs-vswitchd.log --pidfile=/var/run/openvswitch/ovs-vswitchd.pid --detach --monitor
We went to check the openvswitch logs but they're empty and contain nothing:
[root@s1-b8 log]# ps -ef | grep -iE 'ovs|openvswitch' root 3369 1 0 May29 ? 00:00:00 ovsdb-server: monitoring pid 3371 (healthy) root 3371 3369 0 May29 ? 00:04:27 ovsdb-server /run/openvswitch/conf.db -vconsole:emer -vsyslog:info -vfile:emer --remote=punix:/var/run/openvswitch/db.sock --private-key=db:Open_vSwitch,SSL,private_key --certificate=db:Open_vSwitch,SSL,certificate --bootstrap-ca-cert=db:Open_vSwitch,SSL,ca_cert --no-chdir --log-file=/var/log/ovsdb-server.log --pidfile=/var/run/openvswitch/ovsdb-server.pid --detach --monitor root 3406 1 0 May29 ? 00:00:00 ovs-vswitchd: monitoring pid 3407 (healthy) root 3407 3406 36 May29 ? 1-00:59:07 ovs-vswitchd unix:/var/run/openvswitch/db.sock -vconsole:emer -vsyslog:info -vfile:emer --mlockall --no-chdir --log-file=/var/log/ovs-vswitchd.log --pidfile=/var/run/openvswitch/ovs-vswitchd.pid --detach --monitor root 3557 1 0 May29 ? 00:01:26 /usr/bin/python /usr/share/openvswitch/scripts/ovs-xapi-sync unix:/var/run/openvswitch/db.sock --log-file
[root@s1-b8 log]# file /var/log/ovs-vswitchd.log /var/log/ovs-vswitchd.log: empty
and because journalctl has been set not to log in XenServer 7 I can't get any logs or metrics from systemd regarding the service (See XSO-543)
This occurs both on XenServer 6.5 SP1 and 7.0