Re: Possible bug: git-svn leaves broken tree in case of error
- From
- Anton Korobeynikov <asl@math.spbu.ru>
- Date
- Oct 31, 2007, 14:04 UTC
- Message-ID
- <1193839479.26720.16.camel@asl.dorms.spbu.ru>
- In-Reply-To
- <20071031084257.GA2911.SS5073SS@mayonaise>
Hello, Eric
> With the following test case, I'm not able to reproduce what you're > describing.
Looks like something nasty here. I also failed to reproduce with such test, however I definitely was sure, that I modelled it properly.
The typical scenario here is that I'm syncing with external repository by hands. I tried to replay this with "bad" authors file, but it doesn't allow me to past through "bad" changeset. And yes, adding new entry to authors files fixes the problem.
I saw corruption, when git-svn in the next run continues fetch changesets. As I saw this problem several times with my current setup, I added some extra backups, so I hope next time I'll catch tree before and after breakage. Sorry for bothering.
Btw, there is way, how repository can be broken (however, by user only): after such error 'git fsck' reports dangling trees, and running 'git gc --prune' will break any future sync.
Anyway, thanks alot.
-- With best regards, Anton Korobeynikov. Faculty of Mathematics & Mechanics, Saint Petersburg State University.