Comment 4 for bug 2016002

I am still getting this behaviour after a charm upgrade (20.03 to 22.03). journal says:

May 11 10:21:14 juju-5b471f-ovn-upgrade-test-13 ovsdb-server[20703]: ovs|00718|stream_ssl|WARN|SSL_accept: system error (Success)
May 11 10:21:14 juju-5b471f-ovn-upgrade-test-13 ovsdb-server[20703]: ovs|00719|jsonrpc|WARN|ssl:127.0.0.1:51952: receive error: Protocol error
May 11 10:21:14 juju-5b471f-ovn-upgrade-test-13 ovsdb-server[20703]: ovs|00720|reconnect|WARN|ssl:127.0.0.1:51952: connection dropped (Protocol error)
May 11 10:21:28 juju-5b471f-ovn-upgrade-test-13 systemd[1]: ovn-ovsdb-server-sb.service: Current command vanished from the unit file, execution of the command list won't be resumed.
May 11 10:21:30 juju-5b471f-ovn-upgrade-test-13 systemd[1]: ovn-ovsdb-server-sb.service: Found left-over process 20703 (ovsdb-server) in control group while starting unit. Ignoring.
May 11 10:21:30 juju-5b471f-ovn-upgrade-test-13 systemd[1]: This usually indicates unclean termination of a previous run, or service implementation deficiencies.
May 11 10:21:30 juju-5b471f-ovn-upgrade-test-13 systemd[1]: Started Open vSwitch database server for OVN Southbound database.
May 11 10:21:30 juju-5b471f-ovn-upgrade-test-13 systemd[1]: ovn-ovsdb-server-sb.service: Succeeded.

systemd says:

# systemctl status ovn-ovsdb-server-sb
● ovn-ovsdb-server-sb.service - Open vSwitch database server for OVN Southbound database
     Loaded: loaded (/lib/systemd/system/ovn-ovsdb-server-sb.service; enabled; vendor preset: enabled)
     Active: inactive (dead) since Thu 2023-05-11 10:21:30 UTC; 31min ago
    Process: 107358 ExecStart=/usr/share/ovn/scripts/ovn-ctl run_sb_ovsdb $OVN_CTL_OPTS (code=exited, status=0/SUCCESS)
    Process: 107503 ExecStop=/usr/share/ovn/scripts/ovn-ctl stop_sb_ovsdb (code=exited, status=0/SUCCESS)
   Main PID: 107358 (code=exited, status=0/SUCCESS)

May 11 10:21:30 juju-5b471f-ovn-upgrade-test-13 systemd[1]: ovn-ovsdb-server-sb.service: Found left-over process 20703 (ovsdb-server) in control group while starting unit. Ignoring.
May 11 10:21:30 juju-5b471f-ovn-upgrade-test-13 systemd[1]: This usually indicates unclean termination of a previous run, or service implementation deficiencies.
May 11 10:21:30 juju-5b471f-ovn-upgrade-test-13 systemd[1]: Started Open vSwitch database server for OVN Southbound database.
May 11 10:21:30 juju-5b471f-ovn-upgrade-test-13 systemd[1]: ovn-ovsdb-server-sb.service: Succeeded.

A manual restart of ovn-ovsdb-server-sb does resolve the issue.

For what its worth I also have the election-timer set high (30s) in this env.