You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by "Peter N. Lundblad" <pe...@famlundblad.se> on 2004/06/05 21:24:28 UTC

Commit problems

Hi,

I was going to commit tonights work and got the following interesting
output. Anyone else seen this before?

Regards,
//Peter

Sending        subversion/libsvn_ra_local/ra_local.h
Sending        subversion/libsvn_ra_local/ra_plugin.c
Transmitting file data ..subversion/libsvn_client/commit.c:747: (apr_err=165001)
svn: Commit failed (details follow):
subversion/libsvn_ra_dav/util.c:670: (apr_err=165001)
svn: MERGE request failed on '/repos/svn/trunk/subversion/libsvn_ra_local'
subversion/libsvn_ra_dav/util.c:367: (apr_err=165001)
svn:
'pre-commit' hook failed with error output:
subversion/libsvn_fs/bdb/bdb-err.c:61: (apr_err=160029)
svn: Berkeley DB error while opening 'copies' table for filesystem /usr/www/repositories/svn/db:
Cannot allocate memory



---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: Commit problems

Posted by "C. Michael Pilato" <cm...@collab.net>.
Ben Collins-Sussman <su...@collab.net> writes:

> On Sun, 2004-06-06 at 02:13, C. Michael Pilato wrote:
> 
> > /me gives sussman the evil eye for not even attempting a catastrophic
> > recovery...
> 
> Huh, I thought that catastrophic recovery required that all logfiles be
> present.  I thought that it was impossible to run if you're using
> DB_LOGFILE_AUTOREMOVE.

Logfiles are only removed when they are no longer needed.  And they
are no longer needed when they hit their size boundary.  So there is
always at least one logfile "live" in the repos.  Further, if some
problem occured which necessitated recovery, a checkpoint is likely to
have been skipped, and that means that all the logs for the current
operation will still be left around (since no checkpoint means no
marking them as "no longer needed").

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: Commit problems

Posted by Ben Collins-Sussman <su...@collab.net>.
On Sun, 2004-06-06 at 02:13, C. Michael Pilato wrote:

> /me gives sussman the evil eye for not even attempting a catastrophic
> recovery...

Huh, I thought that catastrophic recovery required that all logfiles be
present.  I thought that it was impossible to run if you're using
DB_LOGFILE_AUTOREMOVE.



---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: Commit problems

Posted by "C. Michael Pilato" <cm...@collab.net>.
Ben Collins-Sussman <su...@collab.net> writes:

> On Sat, 2004-06-05 at 19:39, Michael W Thelen wrote:
> 
> > $ svn up
> > svn: PROPFIND request failed on '/repos/svn/trunk'
> > svn:
> > Could not open the requested SVN filesystem
> 
> Dunno why it was down, this looks like what happen last week.  I wasn't
> able to run recovery on the repos either.  I ended up restoring the
> backup of the r9929 repository, which is now online and fine again.

I, however, *was* able to restore the busted repository that sussman
squirreled away for examination, and r9930 was safely tucked inside
it.  

/me gives sussman the evil eye for not even attempting a catastrophic
recovery...

I see that Garrett recommitted 9930 anyway -- thanks, Garrett.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: Commit problems

Posted by Garrett Rooney <ro...@electricjellyfish.net>.
Ben Collins-Sussman wrote:
> On Sat, 2004-06-05 at 19:39, Michael W Thelen wrote:
> 
> 
>>$ svn up
>>svn: PROPFIND request failed on '/repos/svn/trunk'
>>svn:
>>Could not open the requested SVN filesystem
> 
> 
> Dunno why it was down, this looks like what happen last week.  I wasn't
> able to run recovery on the repos either.  I ended up restoring the
> backup of the r9929 repository, which is now online and fine again.

That's comforting...

> Who committed r9930?  Can they recommit?

That was me.  I'll see what I can do.

-garrett

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: Commit problems

Posted by Ben Collins-Sussman <su...@collab.net>.
On Sat, 2004-06-05 at 19:39, Michael W Thelen wrote:

> $ svn up
> svn: PROPFIND request failed on '/repos/svn/trunk'
> svn:
> Could not open the requested SVN filesystem

Dunno why it was down, this looks like what happen last week.  I wasn't
able to run recovery on the repos either.  I ended up restoring the
backup of the r9929 repository, which is now online and fine again.

Who committed r9930?  Can they recommit?




---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: Commit problems

Posted by Michael W Thelen <th...@cs.utah.edu>.
* Peter N. Lundblad <pe...@famlundblad.se> [2004-06-05 15:12]:
> I was going to commit tonights work and got the following interesting
> output. Anyone else seen this before?

I'm getting problems accessing the repository at all (same through the
Web interface and ViewCVS, of course).

$ svn up
svn: PROPFIND request failed on '/repos/svn/trunk'
svn:
Could not open the requested SVN filesystem

-- Mike

-- 
Michael W. Thelen
Never express yourself more clearly than you are able to think.
                -- Niels Bohr