30 May
2015
30 May
'15
2:41 p.m.
Hi all, FYI the weirdness from last night (11pm) until this afternoon (2pm) is explained as follows: There was a power outage at our colo facility (HE). After our server came back up when the power returned, it failed to detect its primary data disk and instead failed over to a backup disk which was still inside the chassis, with a full dataset from 2012 before we upgraded. As services booted they resumed operation with old data. When I was able to log in this morning I unmounted that disk, cleaned the filesystems and remounted the correct disk, so we should be operating with correct data now.