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

Re: FW: svn .21.0 problem

From: Ben Collins-Sussman <sussman_at_collab.net>
Date: 2003-06-10 17:17:58 CEST

Jani Averbach <jaa@cc.jyu.fi> writes:

> On Tue, 10 Jun 2003, Ales Fireder wrote:
>
> > > does anybody knows what is causing the following error:
> > >
> > > C:\>svn co http://lalai.hermes.si/repository
> > > svn: RA layer request failed
> > > svn: The path was not part of a repository
> > > svn: PROPFIND of /: 405 Method Not Allowed
>
>
> What does your apache's error log say?
>
> I guess that you have a permission problem.
>
> http://www.contactor.se/~dast/svn/archive-2003-03/0174.shtml
>

I've noticed a real common pattern here. Whenever httpd has a problem
opening the berkeley environment (because of a permissions problem),
the user always sees a "PROPFIND on /: 405 Not Allowed" error from the
svn client. It isn't until one looks at the apache errorlog that one
sees the "true" libsvn_fs permission error.

This has something to do with the way ra_dav works: if the initial
PROPFIND on the "deep" path fails, it tries chopping off a component,
and tries again. Eventually it tries a doing a propfind on '/', and
when that fails, throws the error to the user.

I don't understand this behavior, and I don't like it either. If
mod_dav_svn generated a real fs error ("can't open berkeley db"), is
this error being marshalled across the network? If not, why not? If
so, why isn't ra_dav throwing it to the user immediately?

I'm tired of such a common problem producing a cryptic error message.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Tue Jun 10 17:20:10 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.