One thing that really sucks on rdiff-backup
I’ve been using rdiff-backup for sometime for my own projects, and I’m actually very happy with the way it works with incremental backups of three of my servers and two desktop computers. But there’s a thing that really sucks: when a connection to a remote host fails, next time it will try to regress all backup changes to previous version. That’s not bad if the backup were actually started, but it does that even when there it didn’t copy a byte from remote host because it was down at the time to stablish the ssh connection, or there was some network issue, or so… I often switch off my internet router before go to sleeping at night, and many times I forget to switch it on again before starting the backup, so this occurs two or three times in a week.
Then, the rdiff-backup script will try to regress all files of every host on next request, that means that it will be running for hours -three or four- checking files just for nothing, before starting to do the backup.
It really sucks. Unfortunatly I’ve not the Python development skills needed to solve this, neither the time to learn it in a reasonable time, so there’s something with I must live. I’ll probably make a workaround on my backup script, so it do not call rdiff-backup if there is not a connection to remote host.
Anyway I want to told you about this issue. Maybe you’ve the same problem, or not at all. X-)
