>>> On Wed, Oct 4, 2006 at 3:00 PM, in message
<452412DC.email@example.com>, "Brian J. Creasy"
> Johnathan Gifford wrote:
>> I'm in the middle of planning an upgrade of our Subversion server
>> 1.3.0 to 1.4.0. While attempting to dump the current repository so
>> can load into a '1.4.0' repository I came across this.
>> svnadmin: Malformed file
>> Okay, got my attention, so did some checking with the book and
>> svnadmin verify. Same error on the same revision number. Not much
>> this has been posted here on the mail list about this issue and I
>> only one link on Google with any hope. That one link did say to
>> revision file from the revprops directory in the db directory of
>> repository. The file was garbage and not clear text. The same holds
>> true for the revs directory.
>> I would roll back each revision, but this revision happened months
>> so that's about 5000 revisions, so this is not very practical and
>> beside, all my incremental dumps with the exception of the last week
>> gone. The repository is fine and fully functional in production.
>> since svnadmin verify failed, this has me wondering how many more
>> revisions are bad after the first one encountered. I do have the
>> notification e- mail indicating what the user did.
>> So what is an admin to do? Any suggestions on how to get my
>> dumped using svnadmin dump? Any suggestions how to fix the garbage
>> revprops and revs file?
> I actually came upon this same exact problem with our main repository
> this morning while doing the same thing you were.
> I think I've found a way to get around our malformed file, but the
> workaround takes advantage of the fact that the revision that is
> is r200 and from about a year ago. HEAD is now at r6700. What we're
> planning on doing is running:
> $ svnadmin dump - r201:HEAD repositoryName > dump.svn
> $ svnadmin create repositoryName2
> $ svnadmin load repositoryName2 < dump.svn
> $ mv repositoryName repositoryName_20061004
> $ mv repositoryName2 repositoryName
> Basically, we're just going to throw out those first 200 revisions
> since, at this stage of development, they don't really matter much at
> all. The only problem I can see with doing it this way is that all
> the revisions will now be skewed by 200 from what they used to be;
> will be r4800. I haven't tested doing the load into a new 1.4
> repository. Only into one of the same version that it is already as
> have not upgraded yet.
> Hope this helps if only a little.
Well, after more digging into our repository, we have one revision with
the 'Malformed file' error and six more within twenty reivsions having a
'Found malformed header in revision file' error. So a copy of our
repository is off to CollabNet for 'hopefull' repair along with copies
of all e-mail notification of those commits describing what transpired
during those revisions. I think my employer is about to find out if the
support contract with CollabNet is worth the money or not.
I was in communication with John Szakmeister about this issue offline.
He and another developer have been trying to track these issues down.
He did suggest that if the offending revision is not referred to in a
later revision and you don't know what happened on that revision, there
is a way to create an 'empty' commit so revision numbers stay in sync.
With that you could dump revision 0:199 and 201:HEAD incrementally.
Import 0:199 into the new repository. Either recreate what transpired
on 200 or load the empty commit, then proceed with loading 201:HEAD.
I'm not sure you're going to be able to dump from revision 201, as it
may need to look into revision 200 for content and still cause an error.
I ran into this earlier, I think. It's been a couple of months since I
tried that, but I didn't know, comprehend, or understand what the
problem was at that time.
Our 'Malformed file' error was due to the repository_name/db/rev/xxxx
file and repository_name/db/revprops/xxxx file were completely garbled.
It looked like binary when displayed unlike understandable text for
other revisions. Double check your rev and revprops file for revision
200. If it is somewhat legible, look for the fsfsverify.py tool that
John Szakmeister has written. It might fix your problem. If it
doesn't, I'll let you (and everyone else) if CollabNet fixes our
revisions our repository.
I think what is interesting about this problem, the repository still
functions. Unless you dump your repository or run svnadmin verify by
chance, you'll never know you have a problem. So the moral of the
story, schedule svnadmin verify to run at least weekly, if not daily.
But just as important make sure you do an incremental dump of each
revision committed and keep them for a certain amount of time so you can
rebuild the repository if the svnadmin verify fails.
To unsubscribe, e-mail: firstname.lastname@example.org
For additional commands, e-mail: email@example.com
Received on Wed Oct 4 22:44:34 2006