Marcel Cox wrote:

I was installing the latest HP Proliant SP when the Abend happened!

Chris

Poison pill is a cluster specific abend and means that because of some
communication issue between he cluster nodes, one node is thrown out of
the cluster to avoid inconsistencies. I suggest you post your problem in
the novell.support.cluster-services forum as this really is a cluster
problem and not an abend problem in the traditional sense.

greg,

Better memory usage? You're using every byte you have.

You have a server with 1GB of RAM (so MEMCALC is not going to do
anything at all for this server sorry), hosting 250GB of storage,
running AppleMac support, a large DS dib set, what looks like a tape
changer unit - so I'm assuming does backups for multiple servers?

Of that 1GB - you peaked at 790MB of NLMs, have had 3GB of protected
mode and java apps being swapping in and out, and currently appear to
have somewhere between 800KB and 100MB of memory available for file
caching.

Please - either put this server out of its misery by shooting it, or get
some MORE memory in there - and the more the better.

As Andrew says, getting nwmkde under control will help, but you are
doing way to much with this server with the memory it has in it.