Sampling Techdocs - Up to February 2014

Here I am again with an overview of interesting TechDocs documents I came across while browsing through the latest publications. If you're completely unfamiliar to Techdocs, here's an introduction to it.

• Technical document : DS8000 Host Ports and Installation Sequence
  Here's a nice document on the "DS8000 host adapters installation sequence, port usage, naming conventions and best practices for host connection". Lots of practical tips.
  

• Presentation : DS8000 HyperPAV presentation and demonstration
  This is kind of a recap but it still contains some practical tips for people starting with HyperPAV. It's a "presentation that explains HyperPAV and compares HyperPAVs to Dynamic and Static PAVs. There is a demonstration that shows the performance benefits of changing from no PAVs, to Dynamic PAVs and ultimately HyperPAVs."

• Presentation : zEnterprise z/OS Usage of SMC-R RoCE
  "SMC-R and RoCE provide high throughput with low overhead communications. This paper is developed to detail the SMC-R and RoCE support in z/OS V2.1." This presentation talks about RDMA and SMC R, SMC R Configuration, Management Commands, CP Usage, Switch Ports, SMC R Details, OSA vs. HiperSockets vs. RoCE ... And I can assure you, not one detail is omitted here.

• White paper : Absolute Capping for zEnterprise LPARs
  Here's an interesting one : "Absolute Capping is a new control which was made available in 2013 supporting the zBC12 and zEC12 GA2 CPCs. It is enabled on zEnterprise via licensed internal code Version 2.12.1 at Driver Level 15 or higher.
  Absolute Capping is available to any LPAR but is of particular interest for non z/OS LPARs such as z/VM and Linux on z that cannot make use of z/OS Workload Manager capacity management functions.
  Unlike traditional LPAR Capping, this enhancement is designed to provide a physical capacity limit enforced as an absolute (versus relative) value that is not affected by changes to the virtual or physical configuration of the system. "

Well that's it for now. And, as I always say : just check them out !

Sampling TechDocs - Up to August 2013

It's been some months, but here I am again with an overview of interesting TechDocs documents I came across while browsing through the latest publications. If you're completely unfamiliar to Techdocs, here's an introduction to it.

• FAQ : SMC-R RoCE FAQ Frequently Asked Question
  Hey, I wish I had that one when I first had to write about this for my introduction on the zBC12. As you've read earlier IBM has announced SMC-R in z/OS v2.1 and RoCE feature support in zEnterprise EC12 and zBC12. SMC-R and RoCE provide high throughput with low overhead communications.For all the rest of the information : see this FAQ.

• Flash : Changes Planned for z/OS Preventive Service Ordering
  "IBM plans a number of changes to z/OS preventive service ordering. z/OS preventive service orders will be based on installed products, which is intended to make order content a better reflection of the systems to be serviced".

• White paper : Couple Data Sets Best Practices and Avoiding Disasters
  Flash : Couple Datasets: Best Practices and Avoiding Disasters
   Two documents on the "cornerstones of your sysplex". The flash is a one-pager summarizing the pitfalls. The white paper has the following topics : Overview, best practice reminders, avoiding disasters, i/o delays, data migration activities and copying or mirroring

• Presentation : An Overview of Mainframe Security for Non-Mainframe Personnel
  "This Power Point provides an introduction and overview to the comprehensive security suite available on System z. The target audience is intended to be IT Management or IT Security personnel who may not be familiar with some of the concepts and facilities of System z".I would add : target audience is definitely also mainframe personnel.

• Technical Document : HyperPAV setup with z/VM and Red Hat Linux on zSeries
  This is a "How to Document on steps required to define DS8000 ECKD DASD Subsystem Disks with HyperPAV feature enabled in a Linux on zSeries environment". Examples are  based on Red Hat 6.3 Linux O/S for zSeries. This is the real technical stuff !

• Technical Document : TS7700 Virtualization Engine Series - VEHSTATS Decoder
  This is a very nice one too, when you have to interpret the VEHSTATS presentation output. It gives you a sample of each possible output and describes how to interpret each field. Essential document !

Well that's it for now. And, as I always say : just check them out !

Sampling Techdocs - September-October 2011

Here I am again with my bi-monthly review of interesting TechDocs documents I came across while browsing through the latest publications.

• FAQ : Running Oracle Database 11gR2 on Linux® on IBM System z™ Servers
  This is a nice addition to the FAQ I referred to last time focusing on theOracle support for Oracle Database 11g Release 2 running on IBM System z. It also talks about upgrade paths, new features and planning for future patch set updates.
  
  
• Flash : DS8000 HyperPAV - Why is there IOSQ time with HyperPAVs
  Interesting flash on HyperPAV, why you can still have IOSQ time and how you can detect it from your RMF reports.
  
  
• Presentation : IBM z/OS + DS8800 Synergy
  This is a very recent prezsentation on System z and DS8800 synergy focusing on three key areas : Performance (zHPF, zWLM and DS8000 I/O priority Manager ...), Business Continuity (HyperPAV, HyperSwap improvements ...) and Management&Growth (1 TB EAVs, Quick Init for CKD Volumes, Dynamic Volume Expansion ...). And ... it comes with the replay of the presentation.
  

• Workshop materials : zEnterprise zManager Workshop
  This is the material from one of the Wildfire Workshops on System z Unified Resource Manager. It's mainly the presentations of course (with or without the speaker notes). Topics covered are : Introduction&Overview, Platform Performance Management and usage scenarios.
  

• FAQ : zHPF
  I think anything you want to know about the High Performance FICON feature is in there : functions, features, prereqs, monitoring . . .
  

• Presentation : TS1140 Tape Drive and Control Unit DFSMSdss Performance results
  This test shows performance results from a "DFSMSdss Full Volume Dump and Restore from disk, with the same amount of data per tape drive, to 3592 E06/C06 vs 3592 E07/C07". It also has a nice 1-page slide comparing the two controllers. The results ? Somewhat predictable ? Yes. The details ? Go and have have a look !
  

Well that's it for now. And, as I always say : just check them out !

PAV vs. HyperPAV

I'm not taking any credit for this post, but if you want a simple and clear explanation of the differences between Static PAV, Dynamic PAV and HyperPAV, just head over to the IBM-Main Discussion list. Bill Fairchild of Rocket Software explains it all in a couple of sentences. I hope he doesn't mind quoting him and if you want some more detail, just get to the thread itself.

"First came static PAVs. You decide where the PAVs are inside each LCU (Logical Control Unit), a collection of up to 256 devices. If you want to change any of them, you have to redo your configuration. Lots of work. Error-prone PITA. Hard to change. You must know your hot devices in advance.

Next came dynamic PAV. The WLM decides within one LCU what should be PAVed for the next WLM interval in order to deliver your requested goals, WLM issues control I/O commands to the controller, PAV array is reset inside the controller, and the new PAV configuration is fixed in concrete (static) until the end of the next WLM interval, maybe 10 minutes later?

Next came HyperPAV. IOS decides on an I/O by I/O basis if a PAV is needed for the next I/O, finds one from a pool of available PAV UCBs, directs the new I/O to a PAVed UCB which the controller knows how to send to the proper device, then IOS returns the PAV UCB to the pool of available PAV UCBs when the I/O ends. You don't have to reconfig. You don't wait until the end of the WLM interval. IOS does no control I/O to tell the controller a new PAV configuration. Instantaneously dynamic as opposed to quasi-static as opposed to seriously static". (Bill Fairchild of Rocket Software)

IBM Announcements : DS8000

As I had a day off yesterday (All Saints' Day) I wasn't able to comment on the new storage announcements. I'm just focusing on the DS8000 in this topic. Last summer, more particularly on 22/08/2006 the new Turbo Models of the DS8000 were announced using the new POWER5+ processors. The announcement included also the new 500GB 7.2K FATA disks and 4Gb Fibre Channel and FICON host adapters. (By the way : the 500GB FATA disks were also announced on the DS6000). Last Tuesday 31/10/2006 additional announcements followed. One of the announcements is the withdrawal of these very models announced in august. The reason is that the DS8000 machine is no longer referred to as the 2107. It is replaced by the 2424 machine type. In addition three machine types are added : 2421, 2422 and 2423. The last digit indicates the warranty period. So this is a major change to the DS8000 : from now one the client can choose the warranty period for the DS8000 instead of having to take the 4-year warranty directly.
Another enhancement on the DS8000 is the HyperPAV feature "planned to be offered on z/OS 1.6 and later releases in November 2006. It "is designed to enable applications to achieve equal or better performance than with the original PAV feature alone while also using the same or fewer z/OS resources.""
What's the difference with dynamic PAV : PAV aliases are pre assigned and Workload manager in z/OS can then dynamically reassign addresses when needed if it sees that some devices need more and others need less. This is mostly reactive, and it takes z/OS cycles to do. With HyperPAV, the aliases are not pre-assigned. Aliases are in a pool and are assigned only when needed. It is done in a more on-demand way. It doesn't use z/OS cycles to reassign Aliases. It does not depend on Workload manager function. It minimizes the number of aliases assigned and therefore, the number of addresses used. The number of PAV-aliases needed may be reduced, taking fewer from the 64k device limitation which is important to large customers. The feature is exclusively offered by IBM and requires purchase of the billable HyperPAV feature and a PAV license on the DS8000 and it also requires z/OS V1.6 and later releases