Re: [rdiff-backup-users] how to shrink/purge/truncate the metadata file (and the file_statistics file)?

[Chris Wilson]

Hi Jeff,

On Sun, 3 Jan 2010, =JeffH wrote:

> My rdiff-backup-data/mirror_metadata.*.snapshot.gz file seems to grow 
> without bound, even though I have perodically used --remove-older-than 
> to remove a number of prior increments. I.e. I expected 
> --remove-older-than to shrink/purge/truncate the metadata file (as well 
> as the file_statistics file), but this seems to not be the case. So 
> eventually my backup disk fills up.

I think there may be a misunderstanding here. Each mirror_metadata file 
contains the metadata for just one backup: either a complete snapshot, or 
in some cases with more recent versions, a patch against the previous one.

The file is not modified in any way after the snapshot is created (afaik), 
and is deleted when the snapshot is deleted. Therefore it cannot "grow 
without bound".

If you mean that each mirror_metadata file is larger than the previous 
one, this must mean that you are backing up more files each time. You 
could try diffing the last two to see why it is growing.

In any case the file should not be huge. For example, one of my backups 
currently covers about 65,000 files, and each mirror_metadata is about 
800kb compressed.

> If not, the only way I know to address the issue is to delete everything 
> on my backup disk when it gets near full, and the start all over with an 
> effectively full initial backup. This seems suboptimal for various 
> reasons.

If it's really a problem then you could just delete the rdiff-backup-data 
directory and backup with --force next time, which removes all older 
increments but does avoid the need to re-transfer all the data.

Cheers, Chris.
--
_____ __     _
\  __/ / ,__(_)_  | Chris Wilson <address@hidden> Cambs UK |
/ (_/ ,\/ _/ /_ \ | Security/C/C++/Java/Ruby/Perl/SQL Developer |
\__/_/_/_//_/___/ | We are GNU : free your mind & your software |