× The internal search function is temporarily non-functional. The current search engine is no longer viable and we are researching alternatives.
As a stop gap measure, we are using Google's custom search engine service.
If you know of an easy to use, open source, search engine ... please contact support@midrange.com.


  • Subject: CPU utilization, Priority, and Throughput
  • From: "Richard Jackson" <richardjackson@xxxxxxxxxxxxxxxxxx>
  • Date: Mon, 17 Jul 2000 08:35:11 -0600
  • Importance: Normal

In the thread "SETI@Home" someone suggested that a batch job running at low
priority adversely effected work running at higher priority.  I am
interested in this issue.

For reasons that I suspect but cannot prove, it seems to me that an AS/400
job running at priority 99 can interfere with jobs running at higher
priorities.  The obvious candidates are:

1. Queued IO that completes after PTY99-job has been preempted by another
task.  I consider RAID parity maintenance and journaling IO (user and
system) to fit into this class.

2. A requirement for vast amounts of memory.  Other jobs must recover
"their" main-storage page frames stolen by the memory-hog job.

3. Some anomaly in priority preempt scheme - perhaps jobs that spend long
unbroken intervals below the MI layer can avoid preemption and annoy
higher-priority jobs

These categories may be distinguished by the effected resource.

I am not concerned that a job running at priority 99 causes one CPU to run
at 100 percent utilization.  As someone said, if the cycles are not being
used by higher priority jobs, that is what it is supposed to do.

I am concerned when the elapsed time for a priority 50 batch job is
significantly increased when a priority 99 job runs and there are no
obviously-overloaded components.

I have four questions:

1. Has anyone else noticed the symptom I described?  In other words, is it
real or am I describing an already-known issue is new terms?  If I didn't
make it up, let us discuss the application to discover if it fits one of the
cause categories.

2. Are there more potential cause categories?  How might they operate to
create the symptom?

3. Is anyone aware of any analysis, modeling, or benchmarking to understand
this phenomenon?

4. Is it sensible to describe causes based on effected resource, is there a
better scheme to describe the issues?

Questions?  Comments?  War stories?

Richard Jackson
mailto:richardjackson@richardjackson.net
www.richardjacksonltd.com
Voice: 1 (303) 808-8058
Fax:   1 (303) 663-4325


+---
| This is the Midrange System Mailing List!
| To submit a new message, send your mail to MIDRANGE-L@midrange.com.
| To subscribe to this list send email to MIDRANGE-L-SUB@midrange.com.
| To unsubscribe from this list send email to MIDRANGE-L-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: david@midrange.com
+---

As an Amazon Associate we earn from qualifying purchases.

This thread ...

Replies:

Follow On AppleNews
Return to Archive home page | Return to MIDRANGE.COM home page

This mailing list archive is Copyright 1997-2024 by midrange.com and David Gibbs as a compilation work. Use of the archive is restricted to research of a business or technical nature. Any other uses are prohibited. Full details are available on our policy page. If you have questions about this, please contact [javascript protected email address].

Operating expenses for this site are earned using the Amazon Associate program and Google Adsense.