Frontend HA RAFT problems

Community Support
21

also what about mac addresses? I setuped very simillar mac addresses to all nodes.

node1 - eth1 - de:6e:be:68:65:8b
node2 - eth1 - de:6e:be:68:65:8d
node3 - eth1 - de:6e:be:68:65:8f

22

As in other thread, we have find out that the default value for
xmlrpc-timeouts are causing this problems. Please change oned.conf to read:

XMLRPC_TIMEOUT_MS    = 0

Cheers

23

hello, thank you! looks like it is working now.

what about this errors in logs? still repeating on followers

Log record 228172 loaded incorrectly. Record index: 228173 fed. index: 0 sql command: . Operation return code: 0

24

This messages could be normal when recovering a node, and may appear in normal operation. You can ignore them by now. We’ll remove them in the next release.

25

ok, what about opennebulla service crashing, today crash on leader, recovery was ok, but availability is lower that one node setup. when will be fix?

26

as a work around you can try to use ONE_AUTH. Set the variable and restart
the service, maybe you can set that in the opennebula unit file.

27

to /var/lib/one/.one/one_auth ?

28

Yes, simple add

....
[Service]
Environment=ONE_AUTH=/var/lib/one/.one/one_auth
...

To the opennebula service file, don’t forget to issue a deamon-reload, and restart the service. Check that the process includes ONE_AUTH in the environment

29

I added, how can I check it?

One more problem, when I execute systemctl stop/retart opennebula them it takes long time and finally systemd kill daemon.

Stopping OpenNebula Cloud Scheduler Daemon...
Aug  2 11:57:59 engine1 systemd: Stopping OpenNebula Cloud Controller Daemon...
Aug  2 11:58:45 engine1 systemd-logind: Removed session 93.
Aug  2 11:59:29 engine1 systemd: opennebula.service stop-sigterm timed out. Killing.
Aug  2 11:59:29 engine1 systemd: opennebula.service: main process exited, code=killed, status=9/KILL
Aug  2 11:59:29 engine1 systemd: opennebula.service: control process exited, code=exited status=2
Aug  2 11:59:29 engine1 systemd: Unit opennebula.service entered failed state.
Aug  2 11:59:29 engine1 systemd: opennebula.service failed.