Well, upon furthur work (restarted the server) it appears that the problem is:

11:02:09 AllocRecord: No space left in volume log.

This appears to be the problem with a singly replicated volume in that the volume log fills up due to being the only replica. I thought we talked about this, but it appears to not have been resolved.

Should the createvol_rep script recognize a singly replicated volume and turn off resolution for them?

Also, why is there so many fixed sizes in coda? Couldn't this log size be a variable size and start generating warnings at some point about being big, but not stop the file server? This kind of failure mode doesn't seem very good for production software.

-- 
Phil Nelson                    NetBSD: http://www.netbsd.org
e-mail: phil@cs.wwu.edu        Coda: http://www.coda.cs.cmu.edu
http://www.cs.wwu.edu/~phil