I have some more info:
The error does not occur when I use ra_local. It could be related to the
issue cited by Karl, I don't know enough about it to comment.
I found the following in my Apache2 error_log.
[Mon Apr 21 11:18:28 2003] [info] (104)Connection reset by peer:
core_output_filter: writing data to the network
[Mon Apr 21 11:18:33 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:18:35 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:18:36 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:18:40 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:18:43 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:18:43 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:19:25 2003] [error] [client 149.21.56.146] Could not
fetch resource information. [301, #0]
[Mon Apr 21 11:19:25 2003] [error] [client 149.21.56.146] (17)File
exists: Requests for a collection must have a trailing slash on the URI.
[301, #0]
[Mon Apr 21 11:19:46 2003] [info] (104)Connection reset by peer:
core_output_filter: writing data to the network
[Mon Apr 21 11:19:51 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:19:55 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:19:56 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:19:59 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:20:02 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
[Mon Apr 21 11:20:02 2003] [info] (32)Broken pipe: core_output_filter:
writing data to the network
-----Original Message-----
From: kfogel@collab.net [ mailto:kfogel@collab.net
<mailto:kfogel@collab.net> ]
Sent: Monday, April 21, 2003 11:19 AM
To: Hensley, Richard
Cc: 'dev@subversion.tigris.org'
Subject: Re: Error in subversion client
"Hensley, Richard" <Richard.Hensley@McKesson.com> writes:
> I always get the following error when I do large (hundreds of files)
> updates, diffs, or merges from my DAV based repository. One of the
> interesting things to note is that large check outs seem to work just
> fine.
>
> svn: RA layer request failed
> svn: REPORT request failed on '/repos/scarab/vendor/current'
> svn: REPORT of '/repos/scarab/vendor/current': Could not read response
> body: An existing connection was forcibly closed by the remote host.
I wonder if this is related to (the same as?) John Goerzen's
http://subversion.tigris.org/issues/show_bug.cgi?id=1224
<http://subversion.tigris.org/issues/show_bug.cgi?id=1224>
Checkouts worked for him as well, but updates were problematic.
Richard, does it behave differently when you do it over ra_svn, or
ra_local?
> I'm using Subversion 0.20.1, and Apache 2.0.44. My server is running
on
> RedHat Linux 7.3, my client is Windows 2000 completely up to date with
> service packs.
>
> I'm not sure where to look for the error message, or what could be
> causing it.
Have you looked in the Apache2 error log?
-Karl
Received on Mon Apr 21 21:47:46 2003