On Aug 16, 2007, at 2:49 PM, Garrett Rooney wrote:
> On 8/8/07, Blair Zajac <email@example.com> wrote:
>> I'm syncing nine repositories from one office to another using the
>> Windows 1.4.4 client from a http:// to a file:/// repository, and I
>> got this on two different occasions:
>> Copied properties for revision 344.
>> Committed revision 345.
>> Copied properties for revision 3svnsync: REPORT request failed on
>> svnsync: REPORT of 'http://hostname/svn/repos': Could not read chunk
>> size: connection was closed by server. (http://pangosrc)
>> Committed revision 346.
>> Copied properties for revision 346.
>> Committed revision 347.
>> Copied properties for revision 347.
>> Committed revision 348.
>> Copied properties for revision 348.
>> The concerning thing is that the message appeared but the sync
>> Is this something to worry about? Was the data successfully
>> transfered over completely? It looks like it was since no successive
>> commits failed.
>> However, to somebody who doesn't know, seeing these warnings doesn't
>> give warm-fuzzies that everything worked correctly.
>> If so, can we have have this message be silenced or give a message
>> saying it retried and worked?
> Since nobody replied and this is still sitting in my inbox...
> Yes, it seems like a problem in svnsync (something somewhere must be
> ignoring an error where it should be breaking out), but it seems
> likely that nothing has gone overly wrong with your repository. If a
> commit had really failed then the next commit would have errored out
> because the revision numbers would be out of sync (i.e. commit to sync
> revision N+1 would have resulted in the creation of revision N, which
> svnsync checks for). Worst case scenario as far as I can see is that
> maybe some revprops weren't copied over, in which case you can just
> recopy them manually with the copy-revprops command.
Thanks Garrett. Good to know that that check is in there.
It would be nice if svnsync could resync all the revision properties,
instead of having to sync each one by hand, as I don't have a list of
revisions which gave an error, but it's not hard to script it either.
To unsubscribe, e-mail: firstname.lastname@example.org
For additional commands, e-mail: email@example.com
Received on Fri Aug 17 08:15:14 2007