I apologize for posting this again but our mail servers have been eating
mail for the last 
month I dont know weather I missed the replies your you never got it. 
I'm still having the same problem with the server. 

 I tried running the server on a FreeBSD box with 16MB of ram but it
thrashed very hard. Anway I need to get it running on the SuSE 6.0
server we have in the lab. 

Every thing is as it was before, I've tryed rebuilding different
versions of the server but still no joy. 

Many thanks Mike


Lincoln M C wrote:
> 
> jaharkes_at_cs.cmu.edu wrote:
> >
> > mclinc_at_essex.ac.uk said:
> > | Error rvm_load_segment returns 201
> > | 16:31:24 rds_load_heap error RVM_EINTERNAL
> >
> > Hmm, let's see, the rvm error reporting is a bit terse, so I'll try
> > to list a couple of things that might be wrong.
> >
> > - Are there any `zombied' codaservers still hanging around?
> 
> No I've kill all the coda processes and restarted them a number of times
> 
> >
> > (a crashed server doesn't just disappear, to avoid losing the
> > rvm-mappings and thread info, they hang around in an endless loop,
> > so they are still hogging the resources that a new server needs.)
> >
> > - Is rvm set up correctly? (is there a file /vice/srv.conf?)
> >   Are you using rvm in files or over raw partitions?
> 
> /vice# more srv.conf
> -rvm /tmp/codalogfile /tmp/codadatafile 330301440
> /vice#
> 
> The files do exsist and they are the about the size I would expect.
> 
> >   Did you use the vice-setup-rvm script, and what were the parameters
> >   you gave?
> 
> I used vice-setup not vice-setup-rvm was this wrong?  I guess vice-setup
> must call vice-setup-rvm.
> 
> I used 8 char auth keys (both the same). The server had a id of 1 . the
> root volume was called 'rootvol', the coda super user had a id of 700
> and a name of codadm.  the Log and data areas where files in /tmp.
> 
> >
> > | I cannot create the route volume as i get:-
> > | /vice/srv#  createvol_rep rootvol E0000100 /vicepa
> >
> > No, you need to have a running server before volumes can be created.
> >
> > Jan
> 
> Thanks for your help.
> 
> Mike
> 
> University of Essex