- Cheryl Watson’s TUNING Letter 2004 No. 2 – z990 Performance vs Expectations
- Typo Correction to TUNING Letter 2004 No. 2
- Important SAS V9.1 Fix
- New APAR for LPARs
1. Cheryl Watson’s TUNING Letter 2004 No. 2 – z990 Performance vs Expectations
The fifty-eight-page 2004 No. 2 TUNING Letter was emailed to electronic subscribers yesterday. The print issues will be mailed within a week. You can purchase a printed copy of the current TUNING Letter for $95 at http://www.watsonwalker.com. Here’s the first page of our new issue:
This is a special issue; one that we’ve been researching for more than six months. The z990 series machines that debuted in September 2003 are the most powerful machines that IBM has ever produced. But not everybody is happy with them. It’s taken us several months of investigation, but we think we understand the reasons behind the apparent underperformance. That’s the focus of this issue; our analysis starts on page 26.
New customized LSPR workload mixes are at the core of this problem. So before we can discuss z990 performance in depth, we need to provide an introduction and update to LSPRs (page 3). It will probably surprise most of our readers to discover their workloads probably don’t match the traditional workload LSPR definitions.
Another factor is the move to faster and fewer processors. Although we wrote about this situation in 1998, it’s more pronounced when moving to the faster z990s. So we’ve updated that article and have included it on page 40.
The 40th anniversary of the IBM S/360 was on April 7, 2004, and it brought a large number of significant announcements. We describe these announcements on page 45 in our What’s New? section.
All of the performance results in this issue were obtained using our BoxScore software product. BoxScore was designed to show the differences in actual versus expected speed and capacity when moving to a new machine or environment. You’ll need BoxScore or something like it to make sure that you got what you paid for.
The next issue is almost complete and will follow this one shortly.
We think this is one of the most valuable issues that we’ve ever produced. We hope you think so too.
2. Typo Correction to TUNING Letter 2004 No. 2
Thanks to Tim Smith from KB Toy Stores for catching a typo on page 52. The next to the last sentence in the section on z/VM 5.1 had an erroneous reference to z/VM 1.5.
3. Important SAS V9.1 Fix
Barry Merrill of Merrill Consultants posted the following flash on his MXG Web site. Please see http://www.mxg.com for more detail. Select Newsletters; then select number Forty Five.
“FLASH: Do NOT use SEQENGINE=V7SEQ, V8SEQ, or V9SEQ under SAS V9.1, until you have installed V9.1 Hot Fix for SAS Note SN-012437, which should be available May 24, 2004. And this is NOT just an MXG/ITRM problem: all SAS datasets built by those SEQENGINEs under V9.1, both those written/copied to tape, or to DASD in sequential format, can be corrupted and unreadable and unrecoverable, with no error message when the dataset was created: ONLY when you attempt to read the data will you discover it is unreadable and the data lost.
“ONLY SEQENGINE=V6SEQ can be used under V9.1 on MVS, without hot fix.”
This is ONLY a problem with SAS V9.1; there are no errors using V8SEQ with SAS V8.2.
The direct link to SAS hot fixes is http://ftp.sas.com/techsup/download/hotfix/hotfix.html.
4. New APAR for LPARs
In this TUNING Letter we discuss again the CPU overhead when there is a high ratio of logical CPs to physical CPs. Greg Dyck, IBM z/OS senior technical staff, points out that a new APAR was released for z/OS 1.4 and z/OS 1.5 to address some of the causes of high system CPU overhead in that situation. See APAR OA05907: Performance Improvement, 28Apr2004.