Re: problems with large commits

Karl Fogel <kfogel@newton.ch.collab.net> writes:

> Justin Erenkrantz <jerenkrantz@apache.org> writes:
> > It's a design flaw in how mod_dav walks the tree - basically, it needs
> > to store the response in a separate pool and use an iterative subpool
> > for each file that it looks up. Right now, we can't use an iterative
> > subpool approach because the response is stored in the same pool as
> > the file lookups. So, the memory growth is linear based on how many
> > files are in a directory.

It's worse than linear.

> >
> > It's somewhere on my todo list, but if someone else wants to tackle
> > it, please be my guest. =) -- justin
>
> Is there an issue for it?

Yes, number 773.

-- 
Philip Martin
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org