TCP/NFS

[Ross Ridge]

In article <532 at comcon.UUCP> tim at comcon.UUCP (Tim Brown) writes:
>Platform:
>	ISC2.2 12M RAM
>	TCP, NFS, X(ISC's)
>	WD8003
>
>What it is doing now is pci complains about not being able to find my
>hostname in /etc/hosts (tho it is there) and the lockd complains about
>something similiar to do with the hostname being wrong.  Now the
>really weird part, if I telnet to one of the other hosts (a 6000)
>using the *internet* address it works but if I try to do it using the
>hostname, it hangs.  Ditto for ping.  If I ping another host using the
>hostname it hangs, but using the internet address it works.  I haven't
>a clue here because right after I installed nfs and tested everything
>I sucessfully mounted filesystems from the other host, mailed messages
>and could telnet in both directions.
>
>Hostname returns the correct hostname.

Assuming you are *not* running /etc/named (the BIND daemon), make
sure /etc/resolv.conf has the one line "nonameserver".  With 
IP, NFS, and X all running you'll also likely want to increase NSTREAM
and NQUEUE in the kernel, you can use "netstat -m" to get idea
of your resource usage.

							Ross Ridge
-- 
Ross Ridge								 //
"The Great HTMU"							[oo]
ross at contact.uucp							-()-
ross at watcsc.waterloo.edu						 //