Thank you for responding Ryan, I already checked permissions on destination repo and they are set properly.
I am not running any command or file within the hook script.
Check this out:
If I mix file and URL access on serverA, where both, repo and mirror are located ....
serverA>svnsync init file:///dest/repo http://serverA/svn/repo
svnsync works fine, BUT If I do the same on different machines (repo on serverA and mirror on serverB)...
serverB>svnsync init file:///dest/repo http://serverA/svn/repo
I get svnsync: REPORT request failed on http://serverA/svn/repo
svnsync: The requested report is unknown.
weird, isnīt ?
Do you know if svnsync was designed to sync repos located on separate machines ?
Thanks a lot
From: Ryan Schmidt [mailto:email@example.com]
Sent: Tuesday, October 24, 2006 6:17 PM
To: Monzalvo, Leopoldo [CIB-IT]
Subject: Re: svnsync URL error
On Oct 24, 2006, at 08:37, Monzalvo, Leopoldo [CIB-IT] wrote:
>>> I am trying to sync my repos using svnsync.
>>> I followed the instructions showed in http://
>>> journal.paul.querna.org/articles/2006/09/14/using-svnsync and
>>> http://svn.collab.net/repos/svn/trunk/notes/svnsync.txt and was
>>> able to sync my repo using the following command:
>>> svnsync init file:///export3/svnrep-bak/repos/svnsync/test file:///
>>> BUT, whenever I use URLs instead of file locations I get the error:
>>> svnsync init http://ny390-sf01a:9050/svn/svnsync-bak/test http://
>>> svnsync: DAV request failed; it's possible that the repository's
>>> pre-revprop-change hook either failed or is non-existent
>>> svnsync: At least one property change failed; repository is
>>> My apache installations works fine since I am able to do normal
>>> repo operations (checkout, commit, update, etc) using URLs.
>>> Does somebody have any clue on what's wrong ?
>> Have you installed a pre-revprop-change hook as described here?
> Yes I did, actually, I could sync my repo using svnsync init
> file:///torepo file:///fromrepo
> But as soon as I use URLs, I got the error message shown below.
Well, hooks work rather differently under file:/// access than under
any other protocol, because with file:/// access there is no server
process, and with the other protocols, there is.
Can the web server user read and execute the hook script?
Are you running any commands or accessing any files within the hook
script? If so, have you remembered to define all executable and file
locations using absolute paths, or else define the PATH variable to
something appropriate beforehand, since when running as a hook, the
path is empty?
To unsubscribe, e-mail: firstname.lastname@example.org
For additional commands, e-mail: email@example.com
Received on Wed Oct 25 15:49:19 2006