Not really getting it how you use the access,
here is my configuration. My Repo is around 3 Gigs with 1800 revisions
running on a mac mini at home.
I usually access parts of the repo via svn+ssh to my machine
(the biggest project being 6000-10000 files, 200 meg or so, dont know it
exactly)
The repo itself is hostet on a machine with direct hd access and the
access is very swift (definitely much faster than dumping files on
sourceforge)
The only thing whic causes a significant slowdown is the artefacting of
files in the .svn dirs which NTFS cannot really handle,
but that is less a problem of subversion, I never really can recall
having a significant 5 minutes checkin delay on this configuration.
In my opinion as the others have statet the problems might be caused
by the network drive access.
However I have never dumped the entire 3 gig repo onto a single dir,
that might slow things down on the client to a crawl, but only on EXT3
and NTFS, which have not a B-Tree like cluster structure, like most
newer filesystems have (well EXT3 has it optionally)
With Reiser, HFS and any other goog filesystem things should be perfect
even at that size.
Jake C wrote:
> I work from home accessing our respoitory over a VPN on a DSL
> connection. Our repository is several thousand files, over 1GB in size.
> Just checking in a single, one-line change takes 5 minutes.
> <br><br>However, I am running DB2 and WebSphere on a 3 Ghz machine with
> only 512MB of RAM. I have another GB of memory being delivered today,
> and I'm hoping that will help. It will certainly help with compile
> time... :-)
Received on Sat Jun 4 06:47:36 2005