Friday, December 30, 2011

OMEGAMON XE for CICS resource usage

Last week I mentioned some performance/resource related APARs for OMEGAMON.  While I'm on the topic I thought I'd mention a technote that came out a few weeks back for OMEGAMON XE For CICS. 

In this technote if you see higher than expected CPU usage in the OMEGAMON CICS XE agent task,  if you have OMEGAMON z/OS,  try using the OMEGAMON z/OS Inspect feature  and see if the module in question using the CPU is KCPWSR0A.  If it is then check out the workaround documented in this technote:

Thursday, December 29, 2011

Upcoming webcast on Tivoli network management and System Automation

In January Kirk Bean, IBM Tivoli Product Manager, will be delivering a webcast on "Improving System Availability with Tivoli Network Management and System Automation".  

In the webcast, Kirk will be discussing such topics as how to speed up problem determination and enhance usability with new network/system/job log integration (i.e. CANZLOG - something I've posted on before), how to reduce diagnostic time and mean time to repair for network problems with enhanced IP packet trace analysis, and how to improve disaster recovery and resource utilization with new high-availability support for GDPS® Active/Active.  GDPS Active/Active, in particular, sounds like a very interesting topic for many customers.

The event will be January 12th at  11 AM Eastern Time.  It sounds like an informative event covering both network management and automation.  The price is right, it's free to attend the webcast. 

Here's a link to sign up:

Thursday, December 22, 2011

AOTC is now part of Pulse

For those with a z/OS focus who also want to get more up to speed with Tivoli, it's good news that AOTC is now part of the Tivoli Pulse conference.  AOTC, Automated Operations Technical Conference, has been an annual event that covers Tivoli with a focus on operational solutions, such as Systems Automation and Tivoli Workload Scheduling.  For those of you not familiar with Pulse, Pulse is the IBM's annual event with a focus on Tivoli solutions.

To be honest, in prior years Pulse has not had a lot of z/OS specific content.  Pulse has been more focused on overall Tivoli concepts, and not so much on z/OS.  With the addition of AOTC content to Pulse as a parallel track, the overall Pulse event has much more robust z/OS coverage.  If you you are interested in both z/OS and Tivoli overall, I suggest that Pulse becomes more of an option.

Pulse is in Las Vegas March 4-7, 2012.  If you are interested in Pulse, here is a link:

For more information on AOTC, here's a link:

Friday, December 16, 2011

OMEGAMON performance related fixes

There are some performance related fixes for OMEGAMON XE for z/OS that you may want to be aware of.

A couple of these relate to OMEGAMON monitoring storage and CSA utilization, and have potential impact in both the 3270 and Tivoli Portal interfaces.  APAR OA37619 addresses cost of data collection and alerting for CUA, Classic and situations in the TEP interface.

APAR OA37647 has impact in the Classic and CUA interface.

APAR  OA36726 addresses high CPU usage in the TEMS address space related to CSA analysis and monitoring.

Another interesting APAR is OA38126.  This APAR relates to high CPU usage in the TEMS address space related to the cost of DASD collection specific to running situation alerts for DASD.  As is often the case, the more data you need to collect, the more the potential overhead.  DASD collection is always a potential issue because many shops will have multiple thousands of UCBs that need to be monitored.  This particular APAR addresses that the DASD Devices workspace will collect information from both RMF and cache data sources.  If a situation alert against this table of data does not require cache information and is able to do what is needed using just RMF data, then the cache data will not be collected.

Tuesday, December 13, 2011

Upcoming webcast on application high availability

Many applications may have components that execute on different hardware and software platforms. This makes it difficult to control and automate each component to improve availability, reduce overall solution risk and decrease operations workload. Would you like to manage both distributed and mainframe environments as one? 

If you are interested in managing application availability in a heterogeneous environment, you may be interested in this upcoming webcast, "Application High Availability in a Heterogeneous Environment".

The event is this Thursday, December 15th at 11 AM Eastern time.  The price is right, the webcast is free.  In you are interested, here is the URL to sign up:

Thursday, December 8, 2011

FYI. I fixed the link on the iPad article

Oops.  I had a bad link for the iPad article.  I went ahead and fixed it.

Here it is again:

Wednesday, December 7, 2011

Mainframe on an iPad

I'd posted a while back about a way to run the Tivoli Enterprise Portal on an iPad.  Here's an article about various alternatives for getting mainframe data on an iPad.  It's an interesting read, and goes through various alternatives from running TN3270 emulation to other options.

Here's a link to the article:

Monday, December 5, 2011

IT's most wanted: Mainframe programmers

Thanks to David Gaertner for pointing me towards this very interesting Infoworld article.

The article points out what many of us already knew:  the mainframe is very much alive and well.   Or as Mark Twain said many years ago, it's death has been "greatly exaggerated".  The core statistic cited in the article:  80 percent of the Fortune 500 companies still use the mainframe.  And let's not forget government at just about all levels, federal, state, and local, still are heavily invested in mainframe technology. 

Why?  The article says it clearly enough.  "The mainframe's staying power in the age of tiny computers is all about its performance for high-volume transactions and its strengths in security and virtualization".   What's the ramification of this?  In the next few years there is the potential for a mainframe skills shortage.  "A Compuware survey of 520 CIOs in large enterprises found that 71 percent are concerned that this looming skills shortage will hurt their businesses...". 

I've had the opportunity to mentor many younger colleagues over the years.  One of the things I often mention to them is the importance of mainframe skills in the coming years.  As the mainframe grey beards (frankly like me), retire in the next 10 to 15 years, the need for mainframe skills will become even more prevalent.

Let's look at it another way.  If every mainframe shop in the Fortune 500 decided tomorrow to switch from the mainframe to other platforms, it would still take many years for that work to get done (and that's really assuming it's even technically feasible in all cases).  Arguably from a business perspective, it's probably more cost effective to identify and train younger talent to close the skills gap. 

If you are interested in the article, here's the link:,0&source=IFWNLE_nlt_stradev_2011-12-01

Thursday, December 1, 2011

Learn about PARMGEN

Yes, the PARMLIB install process I've mentioned on earlier blog posts has been renamed to PARMGEN.  That makes sense since the term Parmlib implies quite a bit of other things to your typical z/OS sysprog. 

If you are new to the PARMGEN process, or just getting started, there was a very good webcast event today that covered PARMGEN in great technical detail (complete with examples).  It's not too late to check out the webcast.  You can still download the material and check out the replay.

Here's the URL:

I recommend you take the time to start getting familiar with PARMGEN.  In future releases of OMEGAMON you can anticipate that the ICAT tool will fade away and that PARMGEN will become the preferred install method.  Don't be shocked when one day,  in a future release, ICAT is gone.

Also, here is a link for the latest technote/documentation on PARMGEN: