Comment 3 for bug 1944014

Well I can't see anything that would obviously cause this, your timezone setting and the refresh.timer seem to imply that snapd should not have started/attempted the refresh so something is clearly wrong. If you could share snapd logs and also possibly turn on SNAPD_DEBUG=1 so if it happens again we can get more verbose logs that would be great, but as-is I don't see any indication as to why this refresh was performing at that time.