[rdiff-backup-users] question about delete and backup..

[Anton Aleksandrov]

Hello,

I am posting my question here, but sorry, if it is being somehow repeated – I wasn’t able to „google” answer.

 

We have average size backup server handling backups from many servers with rdiff-backup. Very nice tool! Usually we run backup every day. Starting point for some servers is mid-day, for some – late afternoon, as we do not want to start backup for 20 servers at once. Then every day we also run „—delete-older-than 8W”. The delete process is being started in the night, when, according to our tests, all backup processes should be already over.

 

I think our users were changing their files a lot recently, because now it is already second time, when „backup” process is overlapping „delete” process. That is actually quite strange – I was thinking, that if any process is accessing rdiff repository, then none of other rdiff processes will be able to do that. Am I wrong? What could cause such overlapping? We are not using „—force” switch, I am 101% sure.

 

All our servers are „linux”, running either 1.2.1 or 1.2.2 version.

 

Thank you for your replies.

 

Regards,

Anton.