Coda File System

Re: Replicated Volume Problem - Getting Desperate now

From: Froody Dude <rod_at_poptel.net>
Date: Fri, 09 Mar 2001 11:45:48 +0000
Hi,

The only update.init script I can find on my FreeBSD box is in the port's
coda-src/scripts/update.init directory.  When this is run nothing happens
'update.init restart'  both  /vice/db/VRDB and VLDB files still remain out of
sync.  Even restarting the update servers on both boxes and finally
restarting both the servers has no effect.

Any help on this would be deeply appreciated.  Also What OS would anyone
recommend with Coda  as our present Coda 5.3.9 with FreeBSD  4.22 is to
unpredictable when setting up and with replication volumes.

Jan Harkes wrote:

> On Thu, Mar 08, 2001 at 12:28:20PM +0000, Froody Dude wrote:
> > I recieved a whole load of messages in the logs and have searched for
> > references but have not been able to resolv the problem.
> >
> > Errors in log on SCM
> >
> > 11:50:16 ViceResolve:Couldnt lock volume 7f000005 at all accessible
> > servers
> > 11:51:39 CheckRetCodes: server nonscm-ip-address returned error 22
> >
> > Errors in log on NONSCM
> >
> > 12:00:48 GetVolObj: VGetVolume(7f000005) error 103
> > 12:00:48 GrabFsObj, GetVolObj error Volume not online
> > 12:00:48 RS_LockAndFetch: Couldnt translate Vid for (0x7f000005.0x1.0x1)
>
> It looks like NONSCM doesn't know that it is hosting that volume.
> Check whether the /vice/db/VRDB and VLDB files have been correctly
> propagated by the updateclnt/updatesrv daemons, i.e. timestamps, size
> and md5 checksums should be identical. If not, restart the update
> daemons on both SCM and NONSCM servers,
>
>     /etc/rc.d/init.d/update.init restart
>
> Jan
Received on 2001-03-09 06:48:07