What is Tivoli Performance Analyzer?

Earlier this week I mentioned that ITM 6.2.3 would be bringing in new feature/function. Tivoli Performance Analyzer in the past has been an add-on component to the ITM infrastructure. With ITM 6.2.3, Tivoli Performance Analyzer becomes part of the base tool.

What is Tivoli Performance Analyzer? At its core, it's intended to add predictive capability to Tivoli Monitoring so you can monitor resource consumption trends, anticipate future performance issues, and avoid or resolve problems more quickly. For example, you can use Tivoli Performance Analyzer to predict application bottlenecks and create alerts for potential service threats.

So, how does it work? Tivoli Performance Analyzer works by analyzing information in the Tivoli Data Warehouse (TDW) and providing an analysis of trends gathered from that data. The user may specify different analysis scenarios, based on what history data is being collected by the user, and the requirements of the user. You have various trending options, such as linear trending or arithmetic trending, and you can use it to calculate such things as time to hit a given threshold.

Tivoli Performance Analyzer works in unison with the TDW, and provides some interesting new analytic capabilities for ITM monitoring solutions (including OMEGAMON). Here's a manual number for more information: SC27-4004-00.

IBM Tivoli Monitoring (aka ITM) 6.2.3 version is now available

It's been a couple weeks now, but ITM 6.2.3 is now available. Don't get ITM 6.2.3 confused with ITM 6.2.2 fixpack 3. ITM 6.2.3 is a new version of the IBM Tivoli Monitoring infrastructure.

What's new in ITM 6.2.3? Quite a few things of interest. Probably the most intriguing and potentially the most powerful is a feature known as self-describing agents. Tired of having to worry so much about application CDs, CAT and ATR files, and looking at ITMSUPER output to identify mismatches in support levels for various agents? Self describing agents makes that process less of a chore. What does it mean? Agents will communicate their application support requirements to the infrastructure dynamically. The infrastructure becomes more flexible and dynamic as a result.

There's a lot more in ITM 6.2.3 that I will be covering over the next few days. To get started on ITM 6.2.3, here's a link:

https://www.ibm.com/developerworks/mydeveloperworks/blogs/0587adbc-8477-431f-8c68-9226adea11ed/entry/ibm_tivoli_monitoring_6_2_3_version_is_now_ega3?lang=en

Interested in capacity planning and System z? Check out this webcast.

If you are interested in System z and capacity planning, there is a webcast next week on September 15th on the topic. Attend "Optimize IT hardware and software expenditures with Tivoli capacity planning and IBM business analytics" to learn about Tivoli “what if” analysis, the Tivoli open architecture for gathering and analyzing data, and application performance modeling and rich graphical reporting.

The webcast will cover the following topics:

• Collecting historical log data on IT utilization to use in performance reporting and usage accounting
• Forecasting and modeling data relationships at the application level
• Using simulation to conduct “what if” analyses that help you avoid unnecessary costs
• Reading, normalizing and storing data for service level optimization
• Delivering business intelligence to be used for predictive analytics

The event is September 15th at 11 AM ET. To attend here is a link:

http://www-01.ibm.com/software/sw-events/teleconference/T805130X92184D07.html

Use ITCAM to identify mainframe as well as distributed bottlenecks

ITCAM (IT Composite Application Monitoring) provides a detailed and dynamic analysis of application transaction flow, and shows end to end application response time, along with aiding in the identification of transaction bottlenecks. What's powerful about ITCAM is it has the ability to 'stitch' together information from a variety of sources to determine the application flow, and show performance information. This end to end perspective could conceivably flow between mainframe and distributed, and encompass various middleware components in between.

The example screenshot shows a distributed client that connects to the mainframe via IMS Connect, and then invokes various IMS transactions. When you look at the example you can see that IMS Connect is the potential bottleneck, and the IMS Connect portion of the response time is highlighted in red on the bottom portion of the screen. This is a fairly straightforward example of how this function operates.

ITCAM is flexible, powerful, and allows you to see where your complex composite applications are spending their time.

Logs logs, where are my logs?

Tivoli distributed and OMEGAMON mainframe kicks out lots of logs and even more messages. These logs can be useful when it comes to problem analysis, and determining why a component may not be functioning as desired.

The logs that are available, and the format of the logs may vary by platform or component. If you want a listing of all the various available logs, here is a link to a handy little document that summarizes where the various logs reside:

https://www-304.ibm.com/support/docview.wss?mynp=OCSSZ8F3&mync=R&uid=swg21502429&myns=swgtiv

Information on z/OS 1.13 OMEGAMON currency maintenance

z/OS 1.13 is coming out, so the question is what, if any, maintenance may be required for OMEGAMON in a z/OS 1.13 environment? The good news is OMEGAMON currency support for z/OS 1.13 is provided for Version 410, 420 and later releases of OMEGAMON XE z/OS products.

Here's a link to a document that lists what, if any, OMEGAMON maintenance is required. Here's the link:

https://www-304.ibm.com/support/docview.wss?mynp=OCSSUSPS&mync=R&uid=swg21512105&myns=swgtiv

By the way, if you are interested in z/OS 1.12 maintenance, here's a link for that, as well:

https://www-304.ibm.com/support/docview.wss?uid=swg21429049

OMEGAMON for z/OS Management Suite V4.2.0 Announced

IBM has announced the OMEGAMON z/OS Management Suite V4.2.0. What exactly is the z/OS Management Suite? Think of it as an integrated offering that includes monitoring for the z/OS operating system, network monitoring, and expanded DASD storage monitoring, all within a single solution.

OMEGAMON XE for z/OS by itself does not really provide much in the way of network information. OMEGAMON XE for z/OS provides DASD, cache, and control unit information, but the available information is largely focused from the perspective of 3390 type devices, etc.

What OMEGAMON z/OS Management Suite V4.2 provides is a way to expand the reach of your core z/OS monitoring to include robust detailed network analysis and more detailed storage information within a single software offering. Technically you could purchase and install OMEGAMON XE for Mainframe Networks and OMEGAMON Storage, in addition to OMEGAMON z/OS. What OMEGAMON z/OS Management Suite V4.2 provides is a way to have all this integrated capability in a single solution.

For more information here is a link:

http://www-01.ibm.com/common/ssi/cgi-bin/ssialias?infotype=an&subtype=ca&htmlfid=897/ENUS211-325&appname=isource&language=enus#h2-abstrx

Interested in cloud?

The IT business has always been full of buzz words and phrases. "Cloud" is certainly one of the latest examples. If you want to learn more about cloud, IBM is putting a series of events that will go into quite a bit of detail of how to get started with highly scalable clouds. The event will look at such things as platform choice for consolidating new cloud workloads and applications, and containing the cost of administering and managing cloud.

If you want to learn more about cloud, the event will be happening at several locations

Fort Worth, Texas September 14, 2011
Markham, Ontario September 20, 2011
Waltham, Massachusetts October 4, 2011
New York, New York October 12, 2011
Chicago, Illinois October 18, 2011
San Francisco, California October 27, 2011
Washington D.C. November 1, 2011

If you are interested, here's a URL to sign up:

ibm.com/software/systemz/seminar/cloud

Data collection considerations for OMEGAMON Storage

There is an interesting technote that discusses collection overhead considerations when doing dataset level monitoring using OMEGAMON Storage. The symptoms may be high CPU in the KCNDLxxx address space, and you may see the following message:

KDFS023A CANDLE DATASET I/O KDFSMIBF ROUTINE IEXP FAILED 0000000C 00000000

What this indicates is that the dataset collection process is overwhelmed with data and is unable to catch up.

This goes back to posts I had made about a year ago relevant to optimal monitoring strategies. In general, the more data you request, the more it will cost to collect and display/manage that data. So consider carefully what you ask for, you may get more than you bargained for.

In this technote the author references the "OMEGAMON XE For Storage Tuning Guide".

http://publib.boulder.ibm.com/infocenter/tivihelp/v15r1/topic/com.ibm.omegamon.stor.doc/SC23-9704-01.pdf

The technote also suggests such things as collecting data every so often (example every 20 I/Os, SAMPCT=20), versus every I/O. Or, collect data based upon an exception criteria, such as a high MSR time for a device.

With so many active UCBs in many shops and potentially so many datasets, it makes sense to have a strategy for collection, versus just turning it on for everything.

Here's a link to the technote:

https://www-304.ibm.com/support/docview.wss?mynp=OCSS2JFP&mync=R&uid=swg21508611&myns=swgtiv

Presentations at Share

At the Share conference in Orlando I did four presentations as part of the proceedings. Here are my topics with links to the material. Each of these presentations contains at least some updates.

Here are the topics and the links to them. Feel free to download:


"Understanding The Interaction Of z/OS Workload Manager And DB2"

http://share.confex.com/share/117/webprogram/Session9395.html


"Monitoring: Become More Proactive Via Effective Historical Analysis"

http://share.confex.com/share/117/webprogram/Session9633.html


"Optimal Alert Management Strategies for System z and Beyond"

http://share.confex.com/share/117/webprogram/Session9776.html


"IMS Performance - Getting The Most Out Of Your Monitoring Technology: Isolating And Solving Common Issues"

http://share.confex.com/share/117/webprogram/Session9808.html

Share presentation on NetView Canzlog

Canzlog is a consolidated log management function that was added as part of NetView V6.1. Canzlog has been popular with quite a few customers as a method of managing multiple log streams in a consolidated manner. It's a nice facility, and worth taking a look at.

Dale Carter of Bank of America) and Tom Howe of IBM did a presentation on "Canzlog in NetView for z/OS 6.1: User Experience".

Here's a link to the presentation:

http://share.confex.com/share/117/webprogram/Session9914.html

Share presentation on the DB2 Catalog

Willie Favero of IBM did an interesting presentation at Share in Orlando on what's new and changed in the DB2 Catalog, "The DB2 for z/OS Catalog Gets a Makeover ". DB2 V10 certainly introduced quite a few changes, and this topic goes through what has happened with the Catalog, what was changed, what was added, and what it means for the DB2 sysprog.

Here's a link for the presentation:

http://share.confex.com/share/117/webprogram/Session9539.html

Upcoming seminars on System z storage

If you are in the Columbus or Charlotte area, you may be interested in a free seminar on how to "Handle common storage management problems with IBM z/OS".

This seminar will look at the Tivoli zStorage Management Initiative. There will be discussion on how Tivoli solutions can reduce time, effort and cost of storage management of your current IBM z/OS storage environment, and reduce costs while fostering greater synergy to address both resource and personnel effectivenesss.

The event will happen September 13 in Columbus, 4499 Fisher Road
Columbus, Ohio, and September 15 in Charlotte, David Taylor Drive
Charlotte, NC.

Registration and Continental Breakfast: 8:30am
Event 9am - 1pm
Lunch will also be served.

The price is right, the event is free (and so is the lunch).

To register - send email to lawatson@us.ibm.com

Overview of z/OS V1.13 at Share

There are quite a few interesting items planned to be coming along in z/OS V1.13. There are quite a few enhancement to z/OSMF (how about web-enabled ISPF), Java/COBOL enhancements, scalability improvements (including performance improvements to of all things IEBCOPY), security enhancements, network improvements, improved availability, more feature/function for WLM and RMF. It's quite a list of items.

If you're interested in learning more about what's coming in z/OS V1.13, here's a link to a nice presentation at Share that goes into it in more detail:

http://share.confex.com/share/117/webprogram/Session9728.html

Information on the z114 at Share

A new z/OS machine is always going to be a hot topic of conversation at Share. The z114 was announced in July, and there multiple sessions that discussed the z114.

The following session is a detailed technical introduction to the z114. The material presented included considerable technical detail on z114 structure, processor design, configuration and features. Here's a link to the presentation:

http://share.confex.com/share/117/webprogram/Session9796.html

A good presentation at Share on managing network performance

Ernie Gilman of IBM did a very nice presentation on "Top 10 Tips for z/OS Network Performance Monitoring with OMEGAMON". This is a very detailed review of common network challenges and how to analyze them using OMEGAMON For Mainframe Networks.

The session looked at how correlation between key metrics can more quickly isolate performance problems, how to identify how poorly designed end to end applications could be the cause to high z/OS overhead, and how to create a simple view to help identify why TCP/IP connections are randomly being rejected.

Ernie made some very nice custom workspaces to show how to more effectively use the tool.

Here's a link to the presentation:

http://share.confex.com/share/117/webprogram/Session9917.html

Notes from the Share Conference in Orlando

I recently returned from the Share Conference last week in Orlando. In addition to presenting on several topics at the conference, I also attended quite a few of the technical sessions (as well as partaking in some hands on labs).

Given that Share has a z/OS focus, many of the sessions were centered around what's new and interesting in the area of System z. That meant there were quite a few sessions on such topics as the new z114 system, what's new with z/OS V1.13, Workload Manager, Unified Resource Manager and zManager, what is zBX and what it means to z/OS, cloud computing, security, storage, you name it.

In the area of zBX, a session I found interesting was done by Glenn Anderson of IBM on "A Mainframe Guy Discovers Blades - as in zEnterprise "Blade" Extension". This was a good introduction to blades and how this technology now folds into the new z/OS mainframes.

Here's a link to the presentation:

http://share.confex.com/share/117/webprogram/Session9674.html

New filter option for Tivoli Data Warehouse history

Thanks to Ernie Gilman for making me aware of this very interesting tidbit.

When you define history collection in Tivoli Data Warehouse (TDW) you go through a set of steps to specify which tables of data to collect, which managed objects to collect the data from, and finally how often to collect the data and send it on to the TDW.

Apparently there a new option has appeared in the TDW history collection options (we think this may have showed up with either FP2 or FP3). When you specify history there is now a new 'Filter' tab. From there you can specify what attribute to filter on at the time of collection. What that means is that you can specify what is essentially exception based history. For example, only collect the history if it meets a specified criteria.

Look at the example I include and you will see how I'm filtering history based on a CPU threshold.

Share Conference

This week I've been at the Share Conference in sometimes sunny, sometimes rainy always hot Orlando. For those of you who may not have heard of Share, Share is a user group with a z/OS focus. Share provides a wide array of sessions on many things related to System z, DB2, IMS, CICS, networking, storage, security, you name it.

I will be doing a series of posts on interesting tidbits of information I picked up a Share.

If you are interested in finding out more about the group go to www.share.org

Seminar on IT optimization

If you happen to be in the Seattle area on August 23rd you may be interested in this seminar on "Optimizing Your IT Infrastructure". The event is intended for Operations managers, infrastructure architects and system programmers.

Here's the agenda

Continental breakfast will be available at 8:30 a.m.
08:40 a.m. - z/OS Version 1 Release 12 and more
09:50 a.m. - Technology update - zEnterprise servers and storage
11:00 a.m. - Consolidation and virtualization update with Linux and z/VM on System z
12:00 p.m. - Lunch
01:00 p.m. - Integrated service management can improve overall visibility into your zEnterprise using OMEGAMON XE for z/OS
02:10 p.m. - Increase availability and productivity in your System z environment with integrated service management automation
03:20 p.m. - Moving to cloud? Integrated service management on System z will be a critical success factor
04:20 p.m. - Close and next steps

It's a full day with some good stuff. It's a free event (so the price is right). If you are interested, here's a link to sign up:

https://www-304.ibm.com/connections/blogs/9051592f-7640-466e-8524-6ae7dcc20c79/entry/seattle_briefing_aug_23_optimizing_your_it_infrastructure?lang=en_us