If any of you have time to answer this, I'm curious about how you do backups. Are you using frequent rsyncs (or something functionally equivalent like BorgBackup), ZFS snapshots, or something custom? It looks like you must have had frequent backups, since when HN finally came back, it was quite close to the state that I remember it being in when it went down.
>>mwcamp+(OP)
Custom, I guess (edit: that was an overstatement—let's say partly custom). We upload a snapshot every 24 hours to S3 and incremental updates every few kb (I think).