FWIW, I do not know if IBM ever fixed it, but TPF (formerly ACP) had no storage key protection between tasks, workloads, and it was very simple, even common, for one task to step on another. If that is still the case, TPF would be the first place I would look for any "strangeness".
Colin Allinson <[email protected]> Sent by: The IBM z/VM Operating System <[email protected]> 10/06/2010 07:58 AM Please respond to The IBM z/VM Operating System <[email protected]> To [email protected] cc Subject Re: Strange performance problem "Quay, Jonathan (IHG)" <[email protected]> wrote :- > Colin, can you say what the workload is? z/TPF guests, VPARS, Linux, CMS? It is a mixed TPF(4.2 - not Z/TPF) guests / VPARS / CMS workload. Naturally, TPF guests are the heaviest users. I have seen this sort of problem before when we have streaming output from one of the TPF guests that has filled all the CP spool or IUCV buffers and I have also seen it when DITTO goes into a loop of retries after running off the end of a tape. I have also seen it when there are huge despatch queues due to very heavy processor load. None of these seem to be the case at the moment. Colin Allinson VM Systems Support Amadeus Data Processing GmbH
