Scyld BeoSetup Hang

[Jarrod Smith]

I had this problem, too.  For some reason, I cannot bring up the slaves
with the CD or I see this behavior.

For me the fix was to make floppies with BeoSetup and use those to boot
the slaves instead of CDs...I never tracked down the difference between
the CD boot and the floppy, but there is something!

YMMV

Jarrod Smith



On Fri, 3 Aug 2001, Sean Dilda wrote:

> On Thu, 02 Aug 2001, Eric T. Miller wrote:
> 
> > Clustering newbie, but not Linux newbie.  Cluster is 6 Celeron500 with
> > supported cards, install Scyld no problem on Master node, check IP
> > connectivity with another Linux box(non-cluster) to ensure NIC's are up.
> > When I boot the first slave node (w/Scyld CD) all goes well up to the RARP
> > requests, which continue indefinatley.  MAC address of slave shows up in
> > BeoSetup, but status is marked as "down".  I assume the node must be in "up"
> > status before continuing with the next step, partitioning the slave from the
> > front end machine.
> 
> You said the MAC address of the slave showed up with a status of 'down',
> so I am assuming you dragged it over to the configured nodes list.  Did
> you also click 'Apply' after dragging it?  If you don't click 'Apply',
> nothing actually gets changed, so the daemons still won't boot the node.
> 

-- 
Jarrod A. Smith
Research Asst. Professor, Biochemistry
Asst. Director, Center for Structural Biology
Computation and Molecular Graphics
Vanderbilt University

jsmith at structbio.vanderbilt.edu