The Question is:
I am experiencing a typical problem on OpenVMS 7.1 running on MicroVAX 4000
system with 64 MB physical memory. The problem is a follows.
I have a application called DMS (Design Management System) that uses Ingres a
backend database. I am experiencing a very high rate of page fault(137
/Transaction) on this particular process, when the same set of commands are
executed on another system wit
h same set of configuration the page fault rate is very minimum (10 Page
Faults/Transaction), could you please tell me why there is a difference ?, I
have tried chaning the system parameters and now it exactly matches with the
system where page falut rate
is low, but still the problem exist and the page fault rate has not come down
?, it will be a great help if you can suggest me how to reduce the page fault
rate on this system
I am giving the system configuration below.
OpenVMS VAX 7.1
The Answer is :
First, determine if there is actually a problem here, and determine
the specific nature of the problem. For instance, on a system
such as this, 137 soft faults is not excessive. (It would also point
to the use of the modified list...) Also, please note that 64 MB is
a relatively small amount of physical memory, and having sufficient
physical memory (for working sets, for I/O caches, and for general
use) can solve a number of performance problems.
It would initially appear that you need more physical memory, or you
need a larger process working set, or a larger WSMAX setting.
Please see the performance management documentation available in the
OpenVMS manual set for a procedure intended to locate the cause of
the problem (excessive faulting), and to resolve it. This includes
use of AUTOGEN, evaluation of process quotas and system parameter
settings, hardware upgrades, and the use of INSTALL and other OpenVMS
tools that can reduce memory requirements.