[OpenAFS-devel] vlserver problems

Jeffrey Hutzelman jhutz@cmu.edu
Wed, 10 Oct 2007 14:43:16 -0400


Please send questions about using and administering OpenAFS to 
openafs-info@openafs.org.  The openafs-devel list is for discussion among 
developers; that is, people actually working on the code.


On Monday, October 08, 2007 09:24:20 PM +0200 Erland Lewin 
<erland@lewin.nu> wrote:

> Hello,
>
> I'm running OpenAFS 1.4.4 on a Linux 2.6.18 server.
>
> I'm having aproblem with my vlserver. In the VLLog, it prints "recovery
> running in state 17" every four minutes. Has some recovery process
> crashed perhaps?

That's a debugging message printed by the ubik recovery thread each time it 
reaches the top of its loop.  Normally this should happen every four 
_seconds_, not every four minutes.  However, the fact that you're seeing 
the message at all suggests that you've turned up the debug level in your 
vlserver.  This would happen if someone sent it SIGTSTP or typed ^Z on its 
controlling tty (normally, the vlserver is run by the bosserver and 
shouldn't have a ctty).


> If I try "vos status localhost", it times out:
>> erl@sol:/afs/lewin.nu/data$ /usr/sbin/vos status localhost
>> Could not access status information about the server
>> Possible communication failure
>> Error in vos status command.
>> Possible communication failure

That suggests you don't have a volserver running.

>  However, I can do a "vos listvldb" successfully.  But not create new
> volumes with "vos create"

So does this, though you haven't said what error you got from 'vos create'.


> I've previously had strange problems with my afs setup related to
> multihomed problems.
>
> Can anyone give me a hint as to how to debug this?

Start by running a fileserver and volserver?

-- Jeffrey T. Hutzelman (N3NHS) <jhutz+@cmu.edu>
   Sr. Research Systems Programmer
   School of Computer Science - Research Computing Facility
   Carnegie Mellon University - Pittsburgh, PA