Application Performance Monitoring is more important than ever as applications grow more and more complicated. The need to: understand the user experience that the application is delivering, auto-discover which infrastructure is delivering that application, monitor transactions end-to-end, perform analytics to understand when the application is not behaving as it normally does, are all needed to ensure these complex applications are available around the clock.
“I see that value of APM solutions for developers and the application teams, but I’m on the infrastructure monitoring team. I have tools that give me much more detail than any APM solution can. Why should I care about APM?”
I get this type of question from the z/OS infrastructure monitoring teams often. You are right, APM is not going to display all of the detailed metrics that the best-of-breed z/OS infrastructure monitoring tools do. z/OS monitoring tools like OMEGAMON collect thousands of performance and availability metrics. These tools help the experts who are responsible for the sub-system health quickly understand the overall health of the sub-systems, and provide the data necessary for deep problem determination.
There are however some real-life benefits for including an APM solution in your end-to-end application monitoring strategy.
- Problem isolation. Give the application teams enough data to figure out if your part of the business application is really the problem or not. With an end-to-end APM solution, isolating where the problem is keeps you out of the war room and lets you sleep through the night when your part of the infrastructure is not the problem.
- Business context. APM tools will give you the needed context to understand where transactions originate to help understand the business impact. If you have to triage multiple problems, knowing which business applications are being impacted will help you make the right decision on which problems to tackle first. This is especially important for junior members of the team. You may know which CICS transactions impact your online banking application, but do the junior engineers?
“Ok, That sounds great, but my application teams already use an APM solution but it doesn’t have any visibility into the mainframe.”
AppDynamics gives a view of the business transaction as it flows across various infrastructure and middleware components. What is missing today is visibility into the Mainframe. Clients we have worked with talk about the Mainframe as a black-box. They can see transactions going into the Mainframe, but they have no visibility into which middleware is involved in the business application, where the transaction is slowing down, or failing.
IBM Z Application Performance Management Connect addresses this critical gap for AppDynamics clients. With Z APM Connect we are able discover where the transaction goes as it enters the mainframe. The solution provides the ability to follow transactions that are currently being monitored by AppDynamics on the open system side into z/OS sub-systems.
Often the same z/OS middleware instances are involved in many different business applications. IBM Z APM Connect can determine which application the transaction originated from and report transaction metrics for that specific transaction instance to the correct business application in AppDynamics. This is done automatically, no manually definition application topologies, or mapping CICS transactions to business applications.
Now that we have automatically discovered the transaction topology and are gathering timing metrics for the transaction segments within z/OS, AppDynamics will perform it’s normal baselining of metrics and automatically notify users if there is a portion of the business transaction that is not performing as it normally does. Automatic detailed snapshots will be captured allowing inspection of where a specific transaction was spending the majority of its time, or where the time was different than normal, to quickly isolate application performance issues.
Our initial release of z Application Performance Management Connect will include support for CICS transaction tracking. We are able to continue tracking transactions from distributed system into CICS via HTTP, CICS Web Services and MQ. Once we have determined a CICS transaction should be tracked, we can continue tracking subsequent CICS transactions in the same or different regions. We will also be able to provide the timing information related to Db2 or IMS DB calls from CICS transactions. The following diagram shows the transaction flows that are supported by Z APM Connect V550. The “Ingress” column shows the entry points into z/OS that are supported from AppDynamics instrumented applications.
Additional application topologies and sub-system types are being prioritized for subsequent development work. If you would like to have your voice heard as far as how this solution evolves, get involved in our Sponsor User program. Or if you would just like to see a demo of the current capabilities contact me. More information about IBM Z APM Connect can be found on our product page ibm.biz/ZAPMConnect as well as on our beta community ibm.biz/ZAPMBeta
If you want to reach out to me please find me on the following channels:
Linkedin – linkedin.com/in/aaron-young-3859295/
Email – mailto:firstname.lastname@example.org
Be sure to subscribe to our bi-monthly zITSM newsletter to stay informed of the latest announcements, upcoming events, educational material, and product introductions and updates. Subscribe at IBM.biz/zSubscribe
Also, check out Doc Buddy v2.0, which now includes top industry news, thought leadership articles, and more.