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

Re: Timeouts and such

From: Ben Collins-Sussman <sussman_at_collab.net>
Date: 2003-05-04 20:51:55 CEST

"Derek J. Balling" <dredd@megacity.org> writes:

> FYI, debug output from my timed out issue is at
> http://www.megacity.org/svnupdate.log

This is very curious. You're running an update, which always begins
with the client sending an http REPORT request, and then describing
its working-copy revisions in the request body.

If the working copy is all at one revision, then the description is
exceedingly short. The more "mixed revisions" the working copy has,
the longer the description.

But your trace is odd -- the body of your request starts out saying
that you have revision 1 of the 'reports' directory. Then it seems to
list *every* single item in the working copy as being at revision 2!
Normally, only the immediate children of your wc-root would need to be
described at revision 2, and that's it, because of implicit recursion.

So my guess as to what's happening: the server is building a mirror
of the working copy, and individually linking *every* object into the
transaction. This is taking a very long time, and the client times
out before the server finishes and can begin calculating a tree-delta.

The real question here is: why the heck is your update-report
describing each item in the wc as a special case? That's the real bug.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sun May 4 20:53:49 2003

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.