Big problem with corosync and pacemaker packages is that they don't put
resources into unmanaged state before upgrade. Hitting an issue like
this is also tightly related to configuration of the pacemaker/corosync.
Upgrading corosync, without stopping pacemaker (and therefore lrmd) has
high probability to result with this behaviour. In the end, corosync is
only a messaging bus.

So the upgrade procedure should really be:
1) put pacemaker's resource into unmanaged mode
2) stop pacemaker
3) upgrade corosync
4) restart corosync
5) start pacemaker
6) put resources into managed mode

Steps 3 & 4 are automatically done by corosync upgrade package, but all
other steps are left to the operator. If upgrade is done by unattended
upgrades, this will quite likely happen again.

-- 
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.
https://bugs.launchpad.net/bugs/1903745

Title:
  upgrade from 1.1.14-2ubuntu1.8 to 1.1.14-2ubuntu1.9 breaks clusters

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/pacemaker/+bug/1903745/+subscriptions

-- 
ubuntu-bugs mailing list
[email protected]
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs

Reply via email to