Re: [rdiff-backup-users] possible memory leak on restore

[dean gaudet]

On Wed, 24 May 2006, Mike Weisenborn wrote:

> I have a fairly good-sized (34G) backup set of which 17G or so is the
> current data, and another 17G is the last 30 days of incrementals.
> 
> I'm trying to create a snapshot of this data in another location by doing
> this:
> 
>    rdiff-backup -r now /target0/backups/host /snapshot/20060523/host
> 
> The problem is, it runs until it gets about 8.4G restored, and then
> memory usage grows from about 90M to about 520M. At that point it
> basically grinds to a halt (swapping).

weird... fwiw i did a ~110GB restore a few months ago... didn't run into 
this problem...

> I ran it with verbosity of 9 and nothing looked out of the ordinary at
> the time when it basically stopped.

is there something odd with the file it was trying to restore at that 
time?

do you have a tmpfs on /tmp or /var/tmp ?

-dean