The other thing to note is that the recover process was killed about 30
mins into the recovery process. I have heard of recover taking far
longer than that on this list. This might be the problem.
Landon
Stuart Robertson wrote:
> Interestingly, I just tried 'strace svnadmin recover /tmp/TestRepos' and it
> completed fine. Now when I run svnlook info /tmp/TestRepos everything works
> fine.
>
> I'm not sure what strace's involvement had, but somehow it's fixed now...?
>
>
>
>>-----Original Message-----
>>From: Tobias Ringström [mailto:tobias@ringstrom.mine.nu]
>>Sent: 20 February 2004 04:32 PM
>>To: Stuart Robertson
>>Cc: users@subversion.tigris.org
>>Subject: Re: Stone-dead repository
>>
>>Stuart Robertson wrote:
>>
>>
>>>Further to my last email, I've just retried "svnadmin lstxns" on the dead
>>>test repository (that is, in /tmp/TestRepos) and svnadmin still hangs.
>>>
>>>
>>
>>If the repository is wedged, you absolutely have to run 'svnadmin
>>recover' before you can access it (using lstxns or anything). You can
>>run 'strace svnadmin recover' to see on a very low level why it hangs.
>>You can post the output here if you want to. To be absolutetely sure
>>that there are no other processes accessing the repository (or a kernel
>>bug), you can always reboot the machine.
>>
>>/Tobias
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@subversion.tigris.org
For additional commands, e-mail: users-help@subversion.tigris.org
Received on Mon Feb 23 22:20:23 2004