ThousandEyes, AppDynamics: New Breed of Application Performance Management
I have yet to come across an enterprise that has used application monitoring in any proactive way. In most cases, admins use systems that use a red light or show an up/down symptom indicator, and look for before they take any kind of action. In essence, we acquired the software but only scratched the surface of most application monitoring software capabilities. More often than not, we never fully configured those tools.
To go even further, you've likely been in a situation where an alert would go off and the up/down indicator was just that - an indicator. It didn't help much, so you had to fish around for the reasons the application was not working properly. Three days of research later, you figured out that the database reached its maximum connection limit.
I am sure some of you will throw in the mix synthetic monitoring and some other fancy keywords. Synthetic monitoring is great, but I still insist that most of us never really configured it or barely got the basic functionality out of it.
Taking all the above into account, we still managed to get by with somewhat acceptable service levels. That was a different time, a different era. It was a time when everything was contained within the boundaries of our data center and we had control over every aspect of the application. A time before SaaS, the cloud, social or mobile.
Today, add all these factors into the mix and you can pretty much render traditional application performance monitoring as obsolete. Now, you have to consider so many new variables, such as your many different SaaS application providers, your cloud provider, the Internet, and, of course your traditional data center. Troubleshooting, monitoring and watching the up/down indicator is no longer a strategy that allows you to just get by in order to maintain any sort of reasonable service levels.
Imagine having an issue and getting into a pointing contest between your internal IT team, your cloud IaaS provider and possibly your SaaS provider -- not to mention your Internet provider -- on where the slowness is, where the outage occurred or who is responsible for fixing it. Imagine getting that call on Monday morning: "Hey, the application was horribly slow on Sunday around 3 pm." Anyone care to troubleshoot that for your CFO with all the factors I mentioned weighing in?
You can see why I am excited about ThousandEyes and AppDynamics. There are other good app monitoring solutions, but those two caught my attention because of their ability to monitor and pinpoint issues within the data center and across the application stack, as well as between the data center and the cloud provider and the Internet provider that the traffic is passing through. You are essentially able to see end to end what is going on with your applications and where you might have impendig issues.
In addition, both companies have significantly aggregated large amounts of data into visually pleasing to navigate dashboards that go beyond traditional up/down monitoring , all the way to exposing the entire landscape of an application with all its interdependencies. IT pros will find themselves actually wanting to navigate through, fully configure and use these tools because of the great value that you can finally derive from them without needing a PhD to configure them.
Another nifty feature: After you've detected an issue you can share and collaborate with coworkers or maybe even professionals from other companies to visually see the problem and work together to address it. The share and collaborate feature is one that will be most valued just because of the ease with which you can share that information. (I know, right? It's such a basic feature, but so powerful.)
Route changes, link delays, bandwidth issues, database connectivity problems, user experience enhancement and more are all new features that these new breed of application performance management tools offer. The exciting thing about some of these companies is they have helped large SaaS providers like Twitter and Citrix GoTo enhance and improve their user experience by detecting potential issues which allows developers or IT professionals to then address by reconfiguring or enhancing the software code.
How many of you are looking at ThousandEyes or AppDynamics today? What has your experience been in a world that is no longer contained within the boundaries of a data center, be it physical or virtual for that matter?
Posted by Elias Khnaser on 03/26/2014 at 2:03 PM