Weheee - hoppy is back on her feet again after her short illness which started on Tuesday. The forum should be in exactly the same state as before she become unwell - there shouldn't be a single second's data loss, as in the end she was brought back to health with a few magical incantations, mainly
e2fsck -y and
e2fsck -cc...the backups from a few hours before her collapse weren't eventually needed. Sincere apologies for the time the forum wasn't available. Most of the downtime was to ensure that the backups were full and safe, just in case the incantations didn't work.
I'm not absolutely certain how she became unwell, but I'm pretty sure it's nothing anyone did on the forum or any hacking attempts. At one point things looked like it was a very bad hard drive problem - there were certainly thousands of "bad blocks" being reported and I'm still unsure whether this is a sign of a possible failing of hoppy's hard drive.
The results of the final fixing incantation
e2fsck -cc seemed to indicate possibly not such a bad problem though, and it may just be that at one point a file was deleted (just in the normal course if things), but the inode (which is a number on the disk which points to the file, something like an index number) wasn't deleted (maybe due to something stopping operations). From what I understand, something like a power outage may be the type of thing that could cause this. Interestingly the day after hoppy became ill, I couldn't connect at all to the server company where hoppy is based in Denmark....I'm wondering if they had any power problems therefore, and whether that may have possibly caused this to happen.
Hopefully this is just an odd random happenining which shouldn't reoccur, and not the sign of a failing hard drive which I'd originally feared. I'll keep a even-more-careful-than-usual watch on hoppy over the next few days and weeks though to ensure she stays happy and healthy.