SUMMARY: Solaris 9 jumpstart problems

From: Adam Kirby (a.kirby@eim.surrey.ac.uk)
Date: Fri Sep 13 2002 - 04:54:22 EDT


Many thanks to Graham Wood and Joshua Symons for their helpful suggestions
which included :

- boot net -v install
  to get greater verbosity in the boot process
- Check for rouge (windows) DHCP servers
  there were none
- check the NFS exports and prehaps share the solaris image and jumpstart
  directory individually. This was not the problem.
- SNOOP, SNOOP and SNOOP again !

Snooped showed something very strange which led to the eventual fix
(although I don't remember seeing this yesterdat) :

client -> server NFS C LOOKUP2 FH=5385 platform
server -> client NFS R LOOKUP2 StaleNFS file handle

So I started to look more closely at the nfs server side. After a bit of
fiddling I up'ed the max number of concurrent NFS requests that the server
can handle from the default 16 to 160 (eg nfsd -a 160) and restarted nfsd
via /etc/init.d/nfs.server. This seemed to fix the problem. This is
strange as the server is not NFS serving anything else.

I say 'seemed' because although it started to work the V880 I was trying
to install started to PANIC with a RED-state on CPU1. It's just sitting
there going mad at the moment :)

Many thanks.

Cheers, Adam

---------- Original message ----------

Managers,

I have a E420R running Solaris 7, setup as a jumpstart server for Solaris
7, 8 and 9. It has been working fine for months but recently I started
having problems....

I'm trying to install a V880 with Solaris 9 - I have installed this
machine before with the same jumpstart server.

When I started this morning I kept getting the infamous 'Timeout waiting
for ARP/RARP packet' despite the fact the mac address is correctly
registered in NIS. I added the mac address to the local ethers file and
the V880 sucessfully completed its tftp download.

I changed the hostname of the V880 and tried again via NIS lookups and it
worked. I've restarted nscd etc...

Anyway the V880 completes it's tftp download, sits there for a while and
then complains that it cannot open it's kernel.

I've run in.rapd and rpc.bootparamd in debug mode and everything looks
correct. I've checked all the obvious things, re-installed the Solaris 9
image etc... but am at a loss.

/etc/bootparams looks fine, jumpstart stuff and solaris image is shared
correctly, tftp is enabled.... I've even rebooted the jump server.

I wouldn't mind but this was working last week :(

Anyone any ideas ??

Cheers, Adam
_______________________________________________
sunmanagers mailing list
sunmanagers@sunmanagers.org
http://www.sunmanagers.org/mailman/listinfo/sunmanagers



This archive was generated by hypermail 2.1.7 : Wed Apr 09 2008 - 23:24:56 EDT