1. About Cheryl Watson’s Tuning Letter 2011 No. 2
2. Analyzing SMF 113 Records
3. Upcoming Conferences
4. zPrime Conclusion
5. IBM Redbooks
6. Authors Wanted
1. About Cheryl Watson’s Tuning Letter 2011 No. 2
The 42-page 2011 No. 2 Tuning Letter was emailed to paid subscribers on May 3rd. You may visit our website at www.watsonwalker.com to obtain subscription infor-mation and the table of contents. The following is the first page from that issue, talking about some of the contents of this latest Tuning Letter:
This issue concludes our three-part series on Parmlib changes from z/OS 1.8-1.12. It took a lot of time and effort, but from the feedback we’re getting, it’s been worth it. Thanks so much for the many flattering comments. If you haven’t taken the time to review your own parmlib members with our recom-mendations, you might want to schedule that time. There will almost cer-tainly be some surprises.
Our next Tuning Letter will cover some of important topics from the last SHARE conference, the promised WLM update, and several reader exchanges.
We have a new z/OS 101 article in this issue that highlights some free training that IBM makes available. There are YouTube videos, webinars, and free workshops. With education budgets so tight these days, here’s a way to learn about newer technologies at no cost.
[Ed. Note – Our z/OS 101 articles are aimed at the new mainframe sysprogs and performance analysts. They are all available to the public on our website by selecting Articles from the index at the left.]
One of the most important enhancements (in my opinion) of z/OS 1.12 is the addition of a new SMF flooding control facility to prevent SMF from filling up and losing records. See more about this new feature on page 21.
When my husband Tom and I first met in 1986, we discovered our mutual love of travel. Since then, we’ve traveled to many parts of the world, including Egypt, India, many Asian countries, Europe (most countries), and South Africa. We’re off to Iran (we planned this before the Middle East protests). You can follow our travels at www.tomandcheryltravels.me.
[Ed. Note – Our vacation to Iran and Turkey was from May 2nd through May 24th, and was a wonderful experience, as you can see from our travel website. What we didn’t expect was that we wouldn’t be able to update our travel journal during the trip. Although internet access was sometimes available, our website was blocked (because it was hosted at GoDaddy), along with Facebook, LinkedIn, Twitter, Blogspot, SHARE, and many other sites.]
2. Analyzing SMF 113 Records
If you’ve already tried, you know that analyzing the SMF type 113 records isn’t the simplest task in the world. These are the Hardware Instrumentation Services records that are available on a z196 (and retroactively on a z10). In Cheryl’s List #147, we described two new tools that we created to show you your Relative Nest Intensity (RNI) and workload (derived from the 113s), which you will need to determine the applicable MIPS for a machine. These are available on our website. (Go to www.watsonwalker.com and select Free Tools on the left.) Of course, if you use IBM’s free capacity tool, zPCR, you can obtain your RNI that way.
In addition to these small tools, there are two other ways you can get some infor-mation from your type 113 records.
- Phoenix Software has recently released zHISR, a tool that uses the type 113 records to identify CPU “hot spots” in your code. Although we haven’t tried it, we assume that this will take a lot less CPU time than most program monitors because it is using hardware collection instead of software collection. It can also use zIIPs for processing. And the price is fairly low. Read more about zHISR at http://www.phoenixsoftware.com/zhisr.htm. What I most like about the tool is that it can automatically start and stop HIS with the proper parameters.
- Peter Enrico, EP Strategies, gives presentations on type 113 records at conferences, and offers to analyze your 113 records at no charge. He will then hold a WebEx session to discuss your results with members of your team, and provide a training session on 113 records. Peter has also given Webinars to train people on type 113s. Contact him at peter.enrico@epstrategies.com. His SHARE session in Anaheim was session 8790 (Exploring the SMF 113 Processor Cache Counters and LSPRs).
3. Upcoming Conferences
You know that we think that conferences provide some of the best education you can receive. There are two upcoming conferences that you might be interested in. (As usual, I’ll be giving a new “Cheryl’s Hot Flashes” session on Friday morning at SHARE in Orlando, although I won’t be able to attend CMG this year. We’ll be in Burma that week.)
SHARE Conference – Orlando, FL – August 7-12, 2011 (www.share.org)
CMG Conference – Washington, DC – December 5-9, 2011 (www.cmg.org)
4. zPrime Conclusion
In Cheryl’s List #148, we discussed a preliminary trial result on the Neon/IBM law-suits. The original lawsuit is finally over, and Neon is withdrawing the zPrime product from marketing. The courts have said that IBM does have the right to restrict the work placed on specialty processors. According to the press, there were no financial penalties made on either side. Here’s a link from the UK’s Channel Register –http://www.channelregister.co.uk/2011/06/01/ibm_prevails_over_neon_zprime/.
Neon issued a press release on May 31 – www.neon.com/neon/news_053111.shtm. Here is the wording of that press release:
NEON SETTLES LEGAL DISPUTE WITH IBM
NEON zPrime to be withdrawn
AUSTIN, Texas – May 31, 2011 – NEON Enterprise Software, LLC today an-nounced it has settled its legal dispute with IBM and will immediately withdraw zPrime from the market. Pursuant to the terms of a permanent injunction, NEON and its distribution partners and affiliates will no longer market, sell, license (including any renewal or extension of any existing license), install, distribute, export, import, offer to sell, offer to license, offer to install, offer to distribute, offer to export or offer to import zPrime.
The U.S. District Court has ruled that (1) only workloads expressly authorized by IBM may be processed on Specialty Engines (including zIIPs and zAAPs) and (2) IBM’s contracts, including the IBM Customer Agreement and the Li-cense Agreement for Machine Code, prohibit software (a) that enables work-loads not expressly authorized by IBM to be processed on Specialty Engines or (b) that circumvents IBM’s technological measures in Machine Code that protect the Built-in Capacity of Specialty Engines and enables workloads not expressly authorized by IBM to be processed on Specialty Engines. Neon has agreed to a permanent injunction under which it will withdraw zPrime from the market and request that licensees and customers remove and destroy their copies of zPrime. Neon will not renew, extend or transfer any existing zPrime license or any warranty, maintenance or service period of any existing zPrime license (or any portion thereof).
Other NEON products are not affected by this settlement.
NEON Enterprise Software, with offices in Sugar Land and Austin, Texas, de-velops and markets software that reduces costs of data center computing. For more information, visit www.neon.com.
I don’t really know if current zPrime customers will recover any money, but I think that it’s a good result for IBM customers because it will retain the usefulness of specialty processors.
5. IBM Redbooks
You probably know what an unabashed fan I am of IBM Redbooks (www.redbooks.ibm.com). These are the first references I use when I want to learn about a new topic or learn the best practices about how to use a facility. These are free publications, and it’s sometimes hard for an IBM department to justify the cost of producing a Redbook. So here’s a way you can help ensure their continuance: When you use a Redbook, please take an extra two minutes to rate the book. There is a clickable link on the right side of each manual description that says Rate this book. Click on the link to rate that particular Redbook, and you can help other readers and help promote the use of Redbooks.
6. Authors Wanted
Watson & Walker is looking for occasional authors for our Tuning Letter. The articles will be similar to those we currently include, but will address areas that would require too much research time for me personally. If you are interested for writing for us, please send a resume and indicate your areas of expertise. If you have already written articles for other publications, please send us an example of your writing. We are especially interested in expertise on DFSMS, GPDS, JES2, sysplex performance, RMF, Unix, Linux, and communications. The articles for our Tuning Letter will need to conform to our standards, such as always including introductory explanations, and providing as many references as possible.
Stay Tuned!