Re: Performance solutions through hardware

We had similar problems with performance.  In fact we dumped ECM because it
was taking too long to generate orders (unfortunately there are only 24
hours in a day!).  Posting of manual journals were timing out and journals
generated from the subsystems were going into an unposted state.  We
experienced many other problems which appeared to be software related but
once we upgraded our processor, the difference it made was significant.
Billing used to take 4 hours and now takes 30 minutes.  This in no way
absolves SSA from their responsibilities in ensuring that the appropriate
indexes are created etc and ECM would still have been a problem as this is
merely poor design (ECM = order entry without the screens). ECM took
approximately 2 minutes to produce on average a 6 line order on a virtually
dedicated machine - totally unacceptable when you generate 700 orders a
day!





fkolmann@revlon.com.au on 99/07/19 07:40:53 AM

Please respond to BPCS-L@midrange.com

To:   BPCS-L@midrange.com
cc:    (bcc: Rosalie Ducasse/ITS/Omnia Group)
Subject:  Re: Performance solutions through hardware




Been on hols for a week, so reply is late.:(

Mike Gillette wrote:

> If in fact the AS/400 is source of your performance bottleneck, it may be
as
> much a workload mix issue as it is a horsepower issue.  The 9406/640 is
> primarily an interactive CPU and suffers under heavy batch processing.
The
> new 7XX models and the older E-Custom Servers are tuned to give the BPCS
> user an adjustable CPW mix between Batch and Interactive.  This has
resulted
> in superior performance than what users have experienced with the
"Classic"
> AS/400 like your 640.

Mike I am very interested in knowing what an Interactive CPU is as opposed
to a
Batch CPU.
Our batch load is not great, or is my definition of BATCH vs INTERACTIVE
something which no longer
applies to the new IBM computers.
I am perplexed as to how hardware can solve a problem that is inherently a
software issue.

To be precise,
if a programmer reads an entire file sequentially to update a record
rather than using a logical file to get to just the record that needs
updating
how will changing the CPU fix the problem.

I am in the dark, I cannot understand how certain jobs take as long as they
do
on the 9406/640.
Certain jobs just scream along, others seem to dissapear up a black hole,
and
occasionally appear and to an update.
Call me paranoid, but this smacks of 'wait loops' to me, not that I would
ever
suggest IBM would ever do anything like
that (does anyone remember a thing call a PCJX or Junior, never mind, its
just a
red herring).
My first job was with a S3M15 so I' ve been about a while and I thought I
knew a
thing or 2 but  this stuff is like
the X-Files to me, or am I paranoid again.  Frank




+---
| This is the BPCS Users Mailing List!
| To submit a new message, send your mail to BPCS-L@midrange.com.
| To subscribe to this list send email to BPCS-L-SUB@midrange.com.
| To unsubscribe from this list send email to BPCS-L-UNSUB@midrange.com.
| Questions should be directed to the list owner: dasmussen@aol.com
+---






+---
| This is the BPCS Users Mailing List!
| To submit a new message, send your mail to BPCS-L@midrange.com.
| To subscribe to this list send email to BPCS-L-SUB@midrange.com.
| To unsubscribe from this list send email to BPCS-L-UNSUB@midrange.com.
| Questions should be directed to the list owner: dasmussen@aol.com
+---