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

Re: Berkeley DB corruption - 'changes' table (again)

From: Paul L. Suh <psuh_at_apple.com>
Date: 2003-10-02 20:35:15 CEST

Karl,

Yes, the admin did turn off httpd before doing the recover.

--Paul

On Oct 2, 2003, at 1:36 PM, kfogel@collab.net wrote:

> When the person ran 'svnadmin recover', did they observe the cautions
> mentioned in
>
> http://subversion.tigris.org/project_faq.html#wedged-repos
>
> ... especially the part about shutting off Apache and any other
> process that might conceivably access the repository?
>
> -Karl
>
> "Paul L. Suh" <psuh@apple.com> writes:
>> Folks,
>>
>> Subversion 0.24.2 (Yes I know it's old, we're planning on updating to
>> 0.30 in 2 days, just as soon as the current project phase
>> finishes. Curse you, Murphy!) We're getting the error:
>>
>> "Berkeley DB error while opening `copies' table for filesystem
>> /Library/Repository/MacOSXTroubleshooting/db: Invalid argument "
>>
>> Three questions:
>>
>> 1) Is there any way to recover? (Not a screaming necessity since we
>> have daily backups, but there were a goodly number of commits
>> yesterday since we're getting close to a release.)
>>
>> 2) Have there been any changes in the BDB table structure in more
>> recent versions that would make this problem unlikely to happen?
>>
>> 3) Is there an Issue Tracker entry for this? I looked but couldn't
>> find one.
>>
>> Additional detail follows.
>>
>>
>> --Paul
>>
>> Paul L. Suh
>> Trainer/Curriculum Developer
>> Apple Computer
>> 1892 Preston White Drive
>> Reston, VA 20191
>>
>> (703) 264-3216
>> http://train.apple.com/
>>
>>
>> Original error e-mail from user:
>>
>>> Hi,
>>>
>>> Any idea how to fix this?
>>>
>>> A little while ago, svn failed due to a full disk on
>>> my machine:
>>>
>>> [da0703a-dhcp110:Current/Production_Materials/Guide]
>>> ellend% svn update 09_FileInternetShareing.fm
>>> Restored 09_FileInternetShareing.fm
>>> subversion/libsvn_wc/log.c:294: (apr_err=155009)
>>> svn: Problem running log
>>> svn: in directory ''
>>> subversion/libsvn_subr/io.c:216: (apr_err=28)
>>> svn: No space left on device
>>> svn: svn_io_copy_file: error copying
>>> '.svn/tmp/text-base/09_FileInternetShareing.fm.svn-base'
>>> to '09_FileInternetShareing.fm.tmp'
>>>
>>> I emptied my trash to free up space.
>>>
>>> Now, when I try to check in a file I get this:
>>>
>>> [da0703a-dhcp110:HDEssentials/Current/Production_Materials]
>>> ellend% svn ci Guide/09_FileInternetShareing.fm
>>> subversion/libsvn_wc/lock.c:116: (apr_err=155004)
>>> svn: Attempted to lock an already-locked dir
>>> svn: working copy locked:
>>> /Users/ellend/Desktop/HDEssentials/Current/Production_Materials/Guide
>>> svn: run 'svn cleanup' to remove locks (type 'svn help
>>> cleanup' for details)
>>>
>>> So then I try to run svn cleanup, and I get this:
>>>
>>> [da0703a-dhcp110:HDEssentials/Current/Production_Materials]
>>> ellend% svn cleanup
>>> subversion/libsvn_wc/log.c:294: (apr_err=155009)
>>> svn: Problem running log
>>> svn: in directory 'Guide'
>>> subversion/libsvn_wc/log.c:1145: (apr_err=155009)
>>> svn: start_handler: error processing command
>>> 'modify-entry' in 'Guide'
>>> subversion/libsvn_wc/log.c:487: (apr_err=155009)
>>> svn: error getting file affected time on
>>> `Guide/.svn/props/09_FileInternetShareing.fm'
>>
>>
>> svnadmin recover produces:
>>
>> Acquiring exclusive lock on repository db.
>> Recovery is running, please stand by...
>> Recovery completed.
>> subversion/libsvn_fs/bdb/bdb-err.c:61: (apr_err=160029)
>> svn: Berkeley DB error
>> svn: Berkeley DB error while opening `copies' table for filesystem
>> /Volumes/data/Repository/MacOSXTroubleshooting/db:
>> Invalid argument
>>
>> Doing a direct db_recover results in:
>>
>> sudo /opt/BerkeleyDB.4.1/bin/db_recover -v
>> db_recover: Finding last valid log LSN: file: 4311 offset 581049
>> db_recover: Recovery starting from [4311][578974]
>> db_recover: Recovery complete at Thu Oct 2 08:30:57 2003
>> db_recover: Maximum transaction ID 8000000b Recovery checkpoint
>> [4311][581049]
>> db_recover: Recovery complete at Thu Oct 2 08:30:57 2003
>> db_recover: Maximum transaction id 80000000 Recovery checkpoint
>> [4311][581049]
>>
>> But svnadmin recover after that still produces the same problem.
>>
>> Systems involved are Mac OS X 10.2.6 on client and server, I built the
>> entire 0.24.2 tarball from source, Berkeley DB 4.1.25, httpd 2.0.46
>> custom built from source, all access via http.
>>
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
>> For additional commands, e-mail: users-help@subversion.tigris.org
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
> For additional commands, e-mail: users-help@subversion.tigris.org
>

---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Thu Oct 2 20:35:53 2003

This is an archived mail posted to the Subversion Users mailing list.

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.