There are quite a few reasons: - If you restart a keepalived service that is acting as MASTER, the floating IP(s) will be removed from that machine, causing service disruption.
- The floating IP(s) may be moved to one of the BACKUP keepalived instances that you may have in the network for a brief moment and then moved back to the MASTER keepalived instance. This produces a sequence of GARP requests that the switches may or may not honor properly. - This can get quite more complicated and disruptive if using scripts that do things like start/stop/restart services when keepalived status changes. Even talking about seconds of downtime are a lot for some services (firewalls, haproxy, VoIP), specially if it is unplanned. - There's always the risk that the service does not start o misbehave after the update, so I definitely prefer to restart it manually during a planned intervention. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/2089155 Title: Do not restart keepalived on unattended-upgrades To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/keepalived/+bug/2089155/+subscriptions -- ubuntu-bugs mailing list [email protected] https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
