so:

spent better part of day trying to figure this one out. after working fine for a while, os x on a LAN failed to mount debian server running nfs-kernel-server. (other debian client worked fine... for reasons that later became obvious.)

restarted everything many times. searched the net as thoroughly as i could. finally found a solution at [site removed, but it's at "andrew's diary"] , peculiarly: putting a hard link to the os x client in / etc / hosts on the server solved the problem, rapidly and totally.

so here's the question: why? why would mountd time-out trying to get back to the client, especially when a pipe is already open? and why did it start failing after working for a while?

any ideas? i hate having a bunch of hacks i'm afraid to play with...

thanks,
.b

I had exactly the same error.

I just added a line with the Mac OS X client IP address to my /etc/hosts like you said, and the mysterious error went away. Like you say, it doesn't make any sense. But it works now!