Joe Williams
New Member
We are running our production databases on OpenVMS v6.2, Progress 7.3c15 and had a perfomance breakthrough increasing the -B from 5,000 to 5,0000. I've determined that the ideal value for this switch for this server is about 330,000. However, I was only able to increase it to 117,000 before I started getting bus errors and the server kept shutting down a few minutes after restart. We are currently running very reliably at 100,000 but would like to achieve optimum throughput and have enough contiguous global pages to theoretically go to 150,000. Once we can break past this mysterious limit, we plan to buy more memory to reach the optimum. I have an open TS call on this but there's not much that can be done from that side at this point. I am planning on getting Compaq involved later this month but, in the interim, has anyone experienced this kind of problem before? Any leads, even on other O/Ss, would be appreciated.