Evangelizing Mainframe
Print Email

Recognizing the Telltale Signs of Poor Application Performance

When my children were younger, they used to call my monitor the “computer.” To them, this was the interface they saw and what went on behind it was not something obvious to them. They were simply naïve and unaware of the complexity making those pictures appear on the screen. It was so cute.
Times have changed, and I find myself occasionally looking at my computer with the same naïve view my children had. That is, I forget I am interacting with a complex arrangement of computers of which I am seeing only the surface level. Recently, I was using a remote desktop to login to a server that was running as a VM guest on a cloud-enabled network. To me, it was simply an image on the screen. I forgot where the “computer” really was, and I’m not sure that I cared (as long as it worked).
While I’m sure you can relate to that, let me conjure up another example for you. You are on the phone with your bank and the representative apologizes saying, “my computer is slow today.” Perhaps, it’s the cable company, insurance agent or an online retailer, but you continually hear the same comment, “my computer is slow today.”
In most cases, I am pretty sure it’s not their computer that’s slow, but the application they’re using. They’re unaware of any details of what is behind the application and what it is doing, but to them, the simple explanation is that “their computer is slow.” And in many cases, that’s good enough.
So what makes it slow? And is it really slow? I am sometimes tempted to say, “Well, when you consider that, while it’s running on a mainframe it’s also referencing 10 databases on five different operating systems and getting confirmation from a third party business partner … it’s actually going really fast, especially compared to the manual process it replaced.”
If you’re the support team in charge of the application, it’s no laughing matter when the application is slow. One of the challenges you face is that it’s probably as complex as I described above, running on a mainframe, multiple servers, across multiple application server instances, combining both legacy, new and third-party applications. So even when you know it’s running slow, knowing where to start looking isn’t obvious.
In many cases, even the team that designed, developed and tested it wouldn’t know where to start analyzing a failure. And typically the support teams responsible for the databases, application servers, and so on, don’t know either. When things are running well, it’s easy to joke that it’s done with magic, but management typically frowns on the “magic” explanation when things are not working.
That’s where application performance monitoring tools come into play, and why they play such an important role in the application ecosystem. Having the ability to see inside of these complex applications and being able to verify when a problem is occurring and where is absolutely crucial for support teams. When the support teams have this level of visibility and insight into the environment and how it really behaves, the business runs smoother and management doesn’t have to understand the complexity or “magic.” When a problem occurs, the time to resolution is reduced without the need to have war rooms and phone conference bridges. More importantly, application performance monitoring can proactively alert these teams and help them resolve issues before they impact the business.
Instead of waiting on the phone with the bank representative, I’d love to hear them say, “My computer sure is fast today!”

Richard Nikula is vice president of development at Nastel Technologies. He has more than 30 years of experience in IT.

Posted: 3/6/2012 9:57:24 AM by Richard Nikula

Print Email

Join Now!