Re: Svn client and multiprocessor machine

D.J. Heap wrote:

Following up with some more information:

I moved Apache and the Subversion repository to a different machine so I
could test just the Subversion client alone on the dual processor
machine. If both processors are enabled, the commands fail as before
within a couple of tries (svn update seems to be more likely to fail).
Disabling a processor once again gets things working without failures.

In addition to the previous test, I also moved my working directory to a
different hard drive (this one an 8GB Fujitsu SCSI on a Adaptec 29160N
controller) on the dual processor machine. Again, if both processors
are enabled, the commands sporadically fail as before (it seems to
succeed slightly more often than it did on the IDE drive, however).
Disabling a processor yields no failures at all after dozens of runs.

So it still appears to me to be some kind of race condition on temp
files in the client only.

The client is single-threaded, so there can't be any race conditions there.

This is really weird, and I have no idea where to start looking. As I
said, my dual WinXP box handles SVN just fine. Let us know if you find
out anything else... perhaps you could try to build the client yourself,
from sources.

-- 
Brane �ibej   brane_at_xbc.nu   http://www.xbc.nu/brane/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org