Sometimes MicroShift failed to run "systemctl restart microshit", return code is 1 and it takes about 30 seconds to return. Then, after a few seconds microshift service is running successfully and pods are in Running state.

In the journal we can see that the shutdown was successful, but during start part an error occurred and MicroShift exited before reporting READY to systemd. This causes systemd/journalctl to inform that the error was "PROTOCOL" which is most likely reference to a "readiness notification protocol" (see https://www.freedesktop.org/software/systemd/man/latest/sd_notify.html)

4.21.0-rc.2

Rare

1.
2.
3.

In these 2 tests:
 - Deploy MicroShift Without LVMS RF
 - Config Flags Are Logged In Audit Flags