Solved

What type of usage drives up Commercial Processing Workload (CPW) in IBM iseries / AS400

Posted on 2011-02-18
8
484 Views
Last Modified: 2012-06-27
I am charged per CPW used during "prime time".   We have robots running, user's in transactions, ODBC queries, native queries, you name it going on.  I don't think IBM publishes the calc, but has anyone tested various types of usage to see what activities use more and less CPW?  Is there something on my dashboard I can monitor to approximate CPW usage?  For example, would a user using a high percent of CPU at a given moment necessarily also be using a high amount of CPW?  Are three users using 1% CPU using the same CPW as one user using 3% CPU?  Any insight on this measurement and how to control it would be greatly appreciated.
0
Comment
Question by:atprato
8 Comments
 
LVL 34

Accepted Solution

by:
Gary Patterson earned 500 total points
ID: 34928694
CPW is just a measure of database-centric system workload that is typical for commercial workloads, and is used typically for comparing the relative "horsepower" of one model of AS/400 to another.  Appendix A explains a little bit about this benchmark:

http://www-03.ibm.com/systems/resources/systems_i_solutions_perfmgmt_pdf_pcrmjan08.pdf

When you say you get "charged" for CPW during prime time, that is usually a charge for CPU utilization, which is one component of CPW.  Also, CPW is also divided into "interactive" (Interactive Feature CPW Utilization) and "batch", and is priced differently since IBM charges more for interactive CPW capacity than "batch" or "server" CPW capabilities.  

Only 5250 workload (interactive users signed onto a green-screen, generally) is charged to Interactive.

You'll need to ask whoever is billing you what specific reports or tools they use to determine the "CPW" that you are getting billed for.  Job accounting data, Performance Collector data, or Management Central collections can be used to gather metrics for charge-back.

So, to answer your question, if CPU utilization is the actual metric that is used to estimate CPW capacity utilization in your environment, then, yes, three users using 1% and 1 user using 3% are the same.

CPW, however, is more complicated than just a measure of CPU, so if a more complex metric is used to figure CPW for chargeback purposes in your environment, then the answer depends on the actual workload being processed, including I/O, database, etc.

- Gary Patterson


0
 

Author Comment

by:atprato
ID: 34930006
Hi Gary, the "if" part confuses me.  Is CPW a moving target?  Is there no standard for what things impact the units of CPW used?  Even if I don't have the exact calculations and I could just use a few key indicators that would be fine.  But you say, "if CPU", can we not even say that CPU is for sure a major component?  Any other measurable major components, and how to measure them?
0
 
LVL 23

Expert Comment

by:Dr. Klahn
ID: 36984688
This question has been classified as abandoned and is closed as part of the Cleanup Program. See the recommendation for more details.
0
Find Ransomware Secrets With All-Source Analysis

Ransomware has become a major concern for organizations; its prevalence has grown due to past successes achieved by threat actors. While each ransomware variant is different, we’ve seen some common tactics and trends used among the authors of the malware.

 
LVL 34

Expert Comment

by:Gary Patterson
ID: 36969876
Oops: didn't see the follow up question.

CPW isn't a "moving target".  

CPW is a general purpose measure of overall system "horsepower".  Like saying a particular gasoline motor generates 275 H.P.  Now, you've got someone that bills you for the "horsepower" that you use when running that motor.  Normal billing for using a motor might be for days, hours, or miles.  How do they convert that to how much horsespower you use?  

I don't know: that should be spelled out in your billing agreement.

System CPW is measured by running a standard IBM-defined workload on given system in a laboratory environment.  As a result, CPW is a fixed number that is associated with a system when it is manufactured.

If you are being charged for the "amount" of CPW you use, then the person doing that billing is using some technique to estimate what percentage of the systems total processing capacity you are using.  They are probably using system job accounting data to do that (google "iSeries job accounting" for more info).  However, job accounting data does not contain a "CPW" measurement, at least not that I know of, so if you are being charged based on "CPW usage", then your vendor is probably using some metric or metrics to estimate the portion of available CPW you are using.

The bottom line is, you'll need to ask them what metrics they are using to perform that calculation, or review the billing agreement to see if it is defined there.

I don't know of any standard mechanism for calculating CPW usage for billing purposes - maybe there is, but I don't know it, and nobody else responded to this thread.

In organizations that bill back system usage, or sell time sharing services, a very common metric is to look at the total number of CPU seconds used versus cumulative CPU seconds available, and use that to estimate usage.  Sometimes usage is charged in multiple rate bands depending on time of day, day of week, etc, with higher rates for peak times, and lower rates for off times.  Sometimes there are additional charges for network bandwidth utilization, disk space usage, memory utilization, number of interactive users, tape usage, optical library usage, etc.

No idea how your provider does it, though.  You really need to ask them.

Hope that helps!

- Gary Patterson
0
 
LVL 27

Expert Comment

by:tliotta
ID: 36970798
I had no reason to respond because there didn't seem to be anything to add. Gary's response is as good as it's likely to get. I don't recall ever hearing of a "billing by CPW" situation before, so I sure don't know what the vendor might be doing. The vendor would seem to be the only source for an actual answer.

Tom
0
 
LVL 34

Expert Comment

by:Gary Patterson
ID: 36970866
I didn't really finish that thought off very well:  

Meant to express: "nobody else responded to the thread, which probably means they don't either."  

Wasn't implying you (or anyone else) should have responded!

- Gary
0
 
LVL 27

Expert Comment

by:tliotta
ID: 36988090
@Gary:

No, no problem. I only intended to echo what you were saying. Sometimes there's no reason to say anything at all; but it seemed useful to add a comment this time, though, just to add another voice in case the OP wasn't sure.

Tom
0

Featured Post

How to improve team productivity

Quip adds documents, spreadsheets, and tasklists to your Slack experience
- Elevate ideas to Quip docs
- Share Quip docs in Slack
- Get notified of changes to your docs
- Available on iOS/Android/Desktop/Web
- Online/Offline

Join & Write a Comment

Stuck in voice control mode on your Amazon Firestick?  Here is how to turn it off!!!
Here's a very brief overview of the methods PRTG Network Monitor (https://www.paessler.com/prtg) offers for monitoring bandwidth, to help you decide which methods you´d like to investigate in more detail.  The methods are covered in more detail in o…
This demo shows you how to set up the containerized NetScaler CPX with NetScaler Management and Analytics System in a non-routable Mesos/Marathon environment for use with Micro-Services applications.

746 members asked questions and received personalized solutions in the past 7 days.

Join the community of 500,000 technology professionals and ask your questions.

Join & Ask a Question

Need Help in Real-Time?

Connect with top rated Experts

9 Experts available now in Live!

Get 1:1 Help Now