Re: [rdiff-backup-users] Repeated index

[Dave Kempe]

Lorenzo Campanelli wrote:
>    Having the "bad filesystem?" clue led me to run fsck.ext3 (without 
> autorepair) a couple of times on the partition, but the error keeps 
> happening.  The rdiff-backup process continues and seems to run 
> indefinitely, although as of this writing I am letting it churn out 
> while it's utilizing 100% CPU (which may be good). Any theories to the 
> cause of the error while I'm waiting?

Can you strace the rdiff-backup process while it is running?
strace -p PIDOFRDIFF-BACKUP
will show you what its trying to do at least.
I have found if you have many many increments (over 1000), then the 
startup time of rdiff-backup gets pretty long, as it stats all the files 
in the rdiff-backup-data directory.

that being said, when I have had corruption of the repository, I have 
found that it can be difficult to recover, and had to start again with 
my increments on more than one occasion :(


>       I tried searching for "repeated index" on the archives and found 
> nothing.  If there is a solution already posted, feel free to bash me as 
> my initiation to the mailing list. 
>     I have tremendous respect for the rdiff-backup solution.  I 
> currently use it for storage-heavy services in a research environment, 
> and would be more than willing to donate to the rdiff-backup project.  
> Are there any project accounts set up to receive funding?

No there is not, but patches are always welcome.
Even posting your stats to the mailing list or starting a wiki page with 
"this is how rdiff-backup works for me" or something is a great way to 
get started contributing.

thanks

dave