'Dear Rocky' answers your System z monitoring and performance tuning questions.
Rocky McMahan, a Senior Software Performance Engineer in z System Software R&D, offers valuable tips to help you get more out of your z System monitoring software.
By streamlining the product and simplifying the architecture, the latest version of OMEGAMON now serves its core customers better while reducing the total cost of ownership in terms of CPU and maintenance.
OMEGAMON has been monitoring IBM z Systems for decades and with the release of the 5.x versions of OMEGAMON, you can see that the product line is strategically important to z Systems software and to IBM overall.
Here is a friendly reminder that you will have an easy test at the end of the article.
An integrated monitoring solution
With the 5.x release of OMEGAMON, customers now have an integrated monitoring solution by leveraging both the Enhanced 3270 User Interface (e3270ui) and the Tivoli Enterprise Portal (TEP) that display vital z/OS and zSeries performance information. This allows subject matter experts or system programmers to promptly respond to issues by using the e3270ui as well as Lines of Business or Operations by using the Tivoli Enterprise Portal, all viewing the same data. Users not only can see z/OS LPAR availability and performance information but with 5.x, they can see this same critical data across their entire sysplex. OMEGAMON provides this data not only for z/OS but also, for CICS, DB2, IMS, MQ Messaging, Storage and Networks. You can also see critical metrics related to your Java Virtual Machines running on z/OS with the recently announced OMEGAMON monitoring for JVM.
Historically the OMEGAMON product set seemed to be developed in silos. In other words, OMEGAMON for CICS would collect a particular metric and communicate that metric differently from the method OMEGAMON for DB2 might use. However, with the 5.x OMEGAMONs, we have standardized on a common platform and are able to share common methods and procedures for doing similar functionality. This results in a simplified architecture that uses fewer resources but at the same time providing more meaningful functionality. This is accomplished but disabling the OMII (CUA) and OMEGAVIEW started tasks and running only one instance (started task) of the e3270ui for the entire enterprise. Customers could see as much as a 48% reduction in CPU usage, when comparing just the cost of running OMII (CUA) and OMEGAVIEW vs. the cost of running the e3270ui. Results at your own shops will vary dependent on your OM II (CUA) and OMEGAVIEW deployments. 5.x also provides the capability to show near-term data for all the OMEGAMONs in the e3270ui. This is the architecture we will build on with future releases of the OMEGAMONs.
That brings us to our first take away from today’s article. Upgrade to the latest OMEGAMON 5.x to take advantage of all the improvements and the most recent architecture. Remember there is a test at the end.
Understand the OMEGAMON architecture and data collection
When it comes to the OMEGAMON architecture itself, there are basically two methods to receive data:
- One method to receive data is via reports. A report can be generated with the TEP, with the e3270ui or with the SOAP interface. Think of a report as a way to send a query to OMEGAMON and ask for metric data from z/OS or a subsystem itself.
- The other method to receive data is via an event or what we call situations. A situation is basically a query to look at values of specific metrics and to tell viewers when some type of condition has occurred. For example, alert me when the CPU usage of a particular address space has crossed a threshold.
OMEGAMON has what we called TEMAs or Tivoli Enterprise Monitoring Agents. The TEMAs are the data collectors for your enterprise. They are installed to gather data from one or multiple systems that are critical to your business (e.g. CICS, DB2, z/OS, MQ, etc.) Data that the TEMA collects is sent to a central repository called the TEMS. The data collected is organized into attribute groups. In addition to collecting the monitoring data for your enterprise monitoring solution, the TEMA can also perform functions, for example,, monitoring for problems on the monitored system, issuing system commands in case a problem occurs, and collecting and storing historical data.
TEMA data collectors organize the data into attributes groups. Attributes are system elements or properties being monitored by an agent, such as CICS Trans ID’s and ASID CPU Usage. These attributes provide a wealth of sysplex- and system-level data. You can use these attributes to tailor the information presented in workspaces, or to define situations that target specific thresholds, events, or performance problems you want to monitor. Think of attributes as fields in a database table and attributes groups as the tables themselves in a database. For OMEGAMON for CICS, there are over 150 attribute groups and each attribute group probably has 20 fields of data. You can now imagine how working together the OMEGAMONs collect a wealth of information about your enterprise.
Learn more about the Tivoli Enterprise Monitoring Server
Now, let’s take a look at the Tivoli Enterprise Monitoring Server or the TEMS. The TEMS consolidates the data collected by the agent (TEMA) and distributes the data either to the TEPS or the e3270ui. It evaluates column functions for situations that cannot be resolved at the TEMA. It sends an alert to the TEPS or e3270ui when specified conditions are met and receives commands from the interfaces then distributes them to the appropriate monitoring agents. The TEMS has an SQL1 Interface as well as a client SQL API which supports a subset of ANSI SQL requests and event/deferred open SQL requests. It also reads metadata catalogs to build internal views of application/table/column information for use in all client requests. Another critical feature is the Situation Monitor, also called Automated Facilities. This operates as a Task Manager to run Situations, Policies and various Housekeeper Tasks.
Why do we explain to you all the detailed capabilities in OMEGAMON? We want you to see the power of this product. OMEGAMON gathers a wealth of metric data and you can see the power that the infrastructure gives you to manipulate the data. By using OMEGAMON, you are just sitting on a rocket! Yes, it gives you an amazing number of capabilities, but you might also get yourself into trouble with this rich feature/function set. Think of it as being similar to giving someone unlimited select/insert/update capability to a database with 5 billion records. That person could write poorly crafted queries that have drastic effects on the database and the machines itself. Our point is, and this is the 2nd take away from the article, with great power comes great responsibility. You have rich capability with the product, but don’t get into trouble with all of that capability.
Now it’s time for the test. Here’s an easy question: what are our two takeaways from this article?
The first is to migrate to OMEGAMON 5.x as quickly as possible and second is that with great power comes great responsibility.