I mentioned yesterday that I received Forrester Report, Market Overview: Application Performance Management, Q4 2011 by Jean-Pierre Garbani. The author talked about APM as anIT Operations journey and specifically he discussed how important it was to understand the process and roles of everyone involved. To me this sounds like he sees some common ground between Applications Performance Management and Business Process Management.
In APM, you’re trying to find the spots where the process breaks down. In BPM you’re defining the process itself, so if your company has an organization around process, and most large organizations are likely to do that, then it might be worth defining your monitoring business process.
He says it’s essential to start with the process and to identify the weakest links, which he believes typically involves alerting and identification, or being aware that a problem actually exists and then nailing down where it is.
Your next step is to map out the roles of everyone involved in that process. If you build a process road map, it makes sense that you should be able to identify who is responsible for each step in the process. This provides a clear list of roles and responsibilities and should help make it easier to understand and identify problems when they happen by checking the map and seeing who should be covering it.
It could also reduce finger-pointing arguments when something goes wrong because it provides a roster of who should be dealing with it.
Garbani believes if you can understand process and roles, you can select the tools that map best to those roles and that can help you when you’re shopping for tools. You can look for ones that match well to the different roles and what each person in your group is trying to accomplish in the system.
While this provides a clearer road map for your monitoring organization, Garbani is careful to point out that there are no “magic bullets” in APM. As he says (and we’ve written here too):
Performance is a function in which all components of your infrastructure, including applications and any form of software, participate. If you don’t monitor everything, you don’t monitor anything, because the unexpected will always happen.
So while you can’t account for every contingency no matter how careful you might plan, you can at least set up a logical system based on process and roles, and at the very least that can bring some more discipline to what can often be a chaotic process.