1. About Cheryl Watson’s Tuning Letter 2009 No. 3
2. What? A Typo?
3. A New Red Alert
4. SHARE in Denver
1. About Cheryl Watson’s Tuning Letter 2009 No. 3
The forty-six page 2009 No. 3 Tuning Letter was emailed to paid subscribers on July 30th. You may visit our Web site at www.watsonwalker.com to obtain subscription information. The following is a summary of just some of the contents of this latest Tuning Letter:
SMF Logger
Two months ago I asked readers to tell me who was running SMF Logger successfully, or if not, what type of problems they had. I received a lot of feedback, and also noticed that there is a great deal of misunderstanding about how to best manage logger. So in this issue, I’ve attempted to gather information from a variety of sources to provide recommendations and things to look for. SMF logger is the best option to keep you from losing valuable SMF data. It also allows you to collect more data than you currently can by using traditional SMF data sets. Not collecting the multitude of DB2 records because of their volume? Then collect those valuable records to help you tune your DB2 system by using logger. If you do any type of chargeback, losing SMF data could result in lost revenue. And with the SMF mechanisms that have been around for years, it’s fairly easy to lose data. Auditors get a little upset about that too. A new update in June from IBM now makes SMF logger a much more manageable feature. All of this does take some system programmer time, but the payback is worth it. For those new to the System Logger, our z/OS 101 section on page 3 provides an introduction.
SMF Management
Of course, that topic led me to continue with my SMF Update series. (As I state in the introduction, SMF data is the basis for all tuning, capacity planning, chargeback, and performance work.) Part 5 shows an update for the BUFSIZMAX parameter of SMFPRMx, and a description of how to use the latest enhancements to SMF logger. I also go into detail for the sysprogs about how SMF buffering works with MANx data sets, and how it changes with SMF logger. Of course, SMF needs to be read, as well as written, so we also show how to efficiently run the two dump programs, IFASMFDP and IFASMFDL. How would you feel about a reduction of 90% in the elapsed time of IFASMFDP?
Elsewhere in this Issue
Our News section provides the normal updates from WSC, important HIPER and New Function APARs, and links to some excellent publications. I’ve added a section to ourNews section to include APARs that affect SMF data. These often aren’t added to normal z/OS maintenance
2. What? A Typo?
We had a couple of minor typos in our Tuning Letter 2009 No. 3 that was just sent out. Our thanks go out to the eagle-eyed readers who told us about them:
On page 2, we said: “Our User Experiences provides a warning about migrating to z/OS 1.10 because the documentation for WLM needs updating.” But nothing was in the newsletter about it. That’s because it’s coming in our Tuning Letter 2009 No. 4, which should be sent out to subscribers next week.
We mentioned several APARs that were marked as ‘FIN’, and I noted that this meant ‘Fixed in Next (or future) Release’. While that was true originally, Jim Horne from Lowes pointed out that now IBM documents this as ‘Fixed IF Next Release’. They aren’t committing, but they’re saying that it’s not important enough for an APAR, but probably more important than a ‘SUG’ (customer suggestion).
3. Red Alerts
IBM issues Red Alerts for especially important APARs, and the most recent one was in early July. You may subscribe to their service at http://www14.software.ibm.com/webapp/set2/sas/f/redAlerts/home.html. In case you don’t subscribe to these alerts, you should at least be aware of them.
2009.07.02 — z/OS 1.8 and z/OS 1.9 with PTFs for OA28473 can result in VSAM datasets with invalid creation dates being deleted. See the explanation and fix in APAR OA29102 (z/OS 1.8–1.9, 2Jul2009) — After Dump / Restore With Replace of KSDS, Original Creation Date of the Restored Dataset is Corrupted. 09/05/26 PTF PECHANGE.
4. SHARE in Denver
It’s that time of year again! We’ll be heading to Denver for the SHARE conference that’s being held August 23rd to August 28th. This conference provides the least expensive training that you can get.
I recently noticed this great offer from SHARE headquarters, and it’s not on their Web site:
50% Discount at SHARE in Denver for those new to z/OS
“Professionals with less than two years experience working on System z are eligible to receive a 50% discount off the full conference rate. When registering, use the code SJCP3 to receive the 50% discount … The 50% discount applies to the regular full-conference registration rate of $2,045. This offer cannot be combined with any other offer. The discount applies to new conference registrations only; no refunds will be given to those who have already registered for SHARE in Denver.”
Academics can appreciate an even greater discount, and students may attend for as little as $250 for the week. I’ve always said that SHARE is the greatest value for your education dollar, and this offer makes it more so! I suggest the track of the zNextGen group. Start with session 2508, a discussion lunch for people new to the Mainframe. For the schedule and complete registration information, please visit www.share.org.
Stay Tuned!