Help! Scyld Beowuld Installation:

[Ao Jiang]

   Hi,
   Thanks a lot for your suggestion.
   When I booted the system, the only 'Fail' item is: Bring up interface
eth1, Determining IP address of eth1... I guess the reason is that the
procotol of eth1 is set as DHCP and active on boot.
   I tried /sbin/service beowulf restart... everything is OK! But my
problem is still exits.
   When I shut down the system, the only 'Fail' item is Starting kill all
/etc/rc.d/rc6.d/S00killall:etc/init.d/beoserv No such file or directory.

   BTW, how much of the new version of scyld beowulf? Is it possible to
get a low price version?

   Thanks agian.

   Tom

On 27 Apr 2002, Sean DIlda wrote:

> On Sat, 2002-04-27 at 21:24, Ao Jiang wrote:
> >   Hi,
> >   I am a beginner of Beowulf user. I tried to install it usingscyld
> > beowulf release 2.0 Preview. I installed the master node successfully. But
>
> Ack! That version's over a year and a half old.. If you are really
> interested in our software, I strongly recommend getting a newer
> version.
>
> > when I installed the slave nodes, I met a problem. According to
> > Installation guide, when I boot the slave node from CD-Rom, the slave
> > node will be listed in beosetup by MAC address. It happened! When I
> > dragged it to center column, clicked 'apply'. The node is assigned
> > to node 0. But the slave node is always 'down', even if I reboot the
> > master node. The status slave node still is 'down'.
> >   By the way, the slave node always send singal "Sending RARP request..."
>
> Based on the message you see on the slave node, I'd say it sounds like
> either the file isn't getting written during the apply, or the daemons
> aren't getting sighup'ed like they should.
>
> My guess is its the daemons not getting sighuped?  Where there any
> problems during boot?  Instead of the green 'OK' message it would have
> had a red 'FAILED' message.
>
> You might want to try forcing the beowulf daemons to restart.  As root,
> do: /sbin/service beowulf restart
> If you see failed messages when its shutting stuff down, don't worry
> about it, however if everything isn't 'OK' when they try to start again,
> then there is a problem there.
>