Our site went down again recently. This was caused by a broken hard drive on the server it's hosted on. It seemed that the company doing the hosting wasn't exactly prepared for it. It seems like they got professional data recovery people in to fix things. But that took ages – and ended in being told that we shouldn't hope for too much. Apparently we hadn't really read the conditions of the service which say that the 'daily backup' only applies to e-mail and SQL databases. Not of much use if you're forwarding all your e-mail anyway and don't do SQL.
In the meantime, unlike our old provider, they were very responsive, though, answered our inquiries within minutes and migrated our site to a different server – meaning no e-mail should have been lost as servers try to re-send stuff for a couple of days.
Our most recent own backup was at the end of May, so it's about a months worth of post that I lost. – And recovered thanks to the help of a couple of my readers who were kind enough to provide their local copies of my feed and Feedster who promised to provide access to my older posts. Thanks a lot everybody. I will re-instate those soon. Comments and trackbacks will unfortunately be lost. I hope things work out in some way.
Another hosting annoyance but at least this time I felt like the provider actually tried to solve our problems and not cheat us. The fact this seemed to be the first time they encountered a problem like this themselves suggests that going with a larger provider may be safer. It may also be more expensive for the same service. And as we don't want to spend too much on this, this may be a compromise we have to live with. – Our provider promised they'll upgrade to some RAID thingy, though, avoiding severe problems caused by hard drive failures in the future.
And I might even make more frequent backups myself.