[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: cvs2svn tests and benchmarks

From: Michael Haggerty <mhagger_at_alum.mit.edu>
Date: 2007-05-12 22:03:22 CEST

Thanks for the info.

We have a separate mailing list for cvs2svn developers, so I set
followups to go to dev@cvs2svn.tigris.org.

Jennifer Starling wrote:
>> 0. Baseline numbers for comparison. If you still have the information
>> from your test conversion, this could probably serve as the baseline.
>> If not, the best baseline would probably be release 1.5.1, using
>> --use-cvs if necessary for your repository.
> command line was:
>
> cvs2svn -v -s /data03/svn-repository/ --branches=branch --tags=tag
> --symbol-default=branch --fallback-encoding=ascii /data02/repository/ >&
> out17 &
> I still have the output log file which is about 763 megs.
>
> I was using version 1.5.1 with the bsd database, running on rhel3u5
>
> I did have to apply the patch posted by Jon Smirl because I kept getting
> KeyErrors. I dont know if that was ever added to the codebase so I
> would have to add it in when using cvs2svn main trunk code I suppose.

The KeyErrors should be gone in the trunk version. So please use the
unpatched trunk version for the benchmarks. If that fails, then we can
think about how to fix it.

> I have plenty of disk space, so I can run it again with the keep files
> option. I'll also check around because I think we have some tools here
> for monitoring memory usage.

Sounds great!

Michael

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat May 12 22:03:49 2007

This is an archived mail posted to the Subversion Dev mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.