Rollback from time mismatch (no hardware clock) before time sync

AFAICS, the state scripts don’t have anything to do with what is happening here. The client is trying to access the network for quite some time, but is failing due to the timesync being missing. The log indicates that this is going on for nine minutes! Is it really not able to get timesync in nine minutes?

You can try to increase the RetryPollIntervalSeconds setting in mender.conf, but to be honest I don’t think it will help. Your best bet will be trying to figure out why the timesync takes so long; it may be broken in some way.

I’ll review what we have in place, and try to check if things have been disrupted in the version transitions.