|Subject:||[rdiff-backup-users] Delete most recent increment?|
|Date:||Mon, 16 Dec 2013 14:57:14 +0000|
|User-agent:||Mozilla/5.0 (Windows; U; Windows NT 6.0; en-GB; rv:18.104.22.168) Gecko/20110414 Lightning/1.0b2 Thunderbird/3.1.10|
I've made a mistake on another server (it holds our library files), which is backed up by riff-backup on our main server, over NFS, each evening. I'd like to delete the most recent increment. How should I do that? (My reading of the manual seems to say the only delete command available is 'remove-older-than', which I do not want to do. I have done a search of the list and I found several messages about deletions, but to do with files, rather than deleting a specific increment.)
In case the context is relevant, here's what happened, but please skip this if not needed:
We have a Library server, and a Main server, running Debian. Their filesystems are each mounted on the other, using NFS. A file library of around 100GB is maintained on the Library server because frequent access is needed, for read purposes, in the normal course of the day; changes to the library are occasional, mostly additions, rarely an edit. The Main server backs up the library files from the Library server, onto its own filesystem, each evening, picking up any library file additions or changes. Rdiff-backup seems very suited to this, and we have been very happy with the scheme.
What went wrong:The Library server was offline at the time of last backup attempt and rdiff-backup backed up an empty library (effectively thinking that all the library had been deleted). The log recorded an increment change of around 100GB (reduction, of course, to zero).
Though this evening the Library server will be online, and rdiff-backup will 'sense' the presence of 100GB of files, it will (now, because of the last increment) treat these as 'new', and therefore do a complete backup again, another 100GB increment. To be honest, this has happened a few times before, but at an additional 100GB backup each time, I will run out of backup space, so I'd like to recover from my mistake by deleting this most-recent increment (of zero files). I am expecting that, if I can delete this unwanted increment, rdiff-backup will continue, as normal but from its previous condition, thus knowing it has a live and current 100GB backup (as, indeed, it has, built incrementally over the years ).
Hence my question. Is there any way I can delete the last increment, so that rdiff-backup performs its increment from its earlier state before it backed up our (then absent) library?
|[Prev in Thread]||Current Thread||[Next in Thread]|