HP3000-L Archives

October 1995, Week 2

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
James Overman <[log in to unmask]>
Reply To:
James Overman <[log in to unmask]>
Date:
Mon, 9 Oct 1995 23:23:01 GMT
Content-Type:
text/plain
Parts/Attachments:
text/plain (34 lines)
Paul H. Christidis ([log in to unmask]) wrote:
: We have an HP3000/967SX running MPE/iX 5.0:
: RELEASE: C.50.02   MPE/iX HP31900 B.79.06   USER VERSION: C.50.02
: We have two applications written in Powerhouse and Business Basic XL.
: They came from a similar machine running MPE/iX 4.x where they shared
: resources with a number of other applications.  We were anticipating
: significant performance improvements with this migration which,
: unfortunately, did *not* materialize.
: Can anyone offer some insights/speculations as to what the problem may be?
 
If the source machine was similar, why were you expecting significant
performance improvements??   Unless the prior system was short on some
resource,  I would not expect a significant change from 4.0 to 5.0.
You do not specify what kind of performance you are measuring,  i.e.
Batch, terminal response, thruput, etc.  But the major areas of difficulties
with 5.0 are 1) not enough memory (a special small memory patch is available),
2) more directories (so wildcard searches can take much longer), 3) a
fragmentation of the disc due to the OS upgrade and patch installation
(either reinstall or try VOLUTIL's CONTIGVOL to condense the discs), 4)
there are some known performance areas on 5.0 (like networks, CM KSAM??,
and some others that have patches available to correct them, see the
Response Center).  Generally, 5.0 on a 967 should be the same or slightly
better than 4.0,  but beware of changes to the number and speed of the discs,
lost tuning such as file placement, TUNE commands, ALLOCATEd programs,  OCT'd
programs,  CPU/Memory differences,  DTC/Lan configurations.
 
If you have quantifiable before and after performance differences, the
HP RC should be able to assist in isolating any OS related performance
bottlenecks.
 
Let us know what you discover as the cause(s) of your system woes.
 
James Overman HP Software Technology Center

ATOM RSS1 RSS2