All back up again finally - at 9:05 UTC

When KDB came back up, no stats were getting to it from ckpool - then ckpool deadlocked itself and crashed.
This of course meant a failover to your backup pools, then fail-back would have started at 9:05 UTC

This cleared up all the remaining problems since both KDB and ckpool had now restarted - and now all is going ok finally.