some things are still broken, and yeah… it’s been fun.
So the disk in my colo el-cheapo box died. No problem, I have a backup right? After the last great disaster. Â Well that disk DIED TOO.
So here we are running on some half baked incremental backup. Â At least I did have this much here we are.
it’s late, I have updates, but Im tired.
Double failures are the worst. Statistically they must happen, but also you can’t justify the cost of backing up backups of backups into infinity. Sorry! Glad you had incrementals.
Yeah it was crazy.. Nothing worse than saying oh well, let me get that 7z, and waiting hours (which felt like days) for it to transfer to find out it has errors.
I even test my backups, but bad disks will be bad disks (the file md5’d the same but I didn’t have a md5 before hand).
As much as I bemoaned the whole trust big companies with my data after that 2-3 week (I forget, but it was a LONG outage a while back), I may end up having to use live drive or whatever it is for another backup destination. I guess it’ll be a good excuse for pgp, and hoping the file doesn’t corrupt.
Consider PAR files to handle any bitrot.
it looks like rsync and the individual files was the win here.. Or at least as far as I can tell, nothing was damaged. There were no errors with the SQL import, which is the one I was really worried about. Although it really would suck to lose all the pictures.
A wise man once said the best backup is to put it on a public FTP site and let it be mirrored.
I myself? Probably just encrypt some compressed VMDKs and send it to some friends.