Anticipations: APM in 2013

You’ll want to read “13 APM Predictions for 2013“. It’s meaty.

While I have deep reservations about the general tradition of end-of-the-year prediction pieces, a few publications make the best of the form. APMdigest is among these, largely because it annually puts together a good line-up of “industry experts”.

Note, especially, the following:

  • For a domain as specialized as application performance management (APM), reliance on experts leads sometimes to a rather incestuous vibe, simply because the universe of experts is so narrow. Most of the article’s thirteen topics have already turned up, sometimes more than once, during the last few months of “Application Monitor“.
  • Most of the readers I’ve met don’t understand thirteen things at a time in any profound way. I know my limit on a good day is four. “You’ll want to read …” at the top of this article is an abbreviation for, “You’ll want to skim … and find the couple of items that matter to you today.”
  • There’s plenty with which I agree. “APM solutions simplified” and “APM SaaS on the rise”, for instance, are especially interesting because they’re not only likely to be true, but also meaningful. Application complexity increases in a lot of IT (information technology) domains, and APM is at an interesting point in its life where simplification is both needed and possible.
  • “APM leverages software defined infrastructure” sounds giddy to me. Of course software-defined infrastructure is coming on strong, as I’ve written and will continue to write in 2013, and of course someone will advertise a successful project that relates APM and more-fully virtualized infrastructure. As a whole, though, the market isn’t ready for sophistication on this score. I expect it to take all year for decision-makers to understand how management of virtualized infrastructure differs from the success of host virtualization (more on this, later). At the end of the year, at least a few vendors will still be working on simplification. With all this going on, there simply won’t be attention left to give to APM’s potential with software-defined infrastructure.
  • The “Convergence of APM and NPM” and DevOps, and crystallization of APM approaches as either Dev or Ops, have been announced already several times. Still, there are nuances that I’ve seen no one explore yet; I hope to return to this topic in 2013.
  • Keep your own interests in mind when reading an article like this. Yes, certain legacy vendors will shuffle off the stage, and yes, “APM competitors co-operate …” is interesting as a dramatic theme, but those realities might not affect your decisions as an APM consumer at all. Is there reason to be optimistic about “Federated APM”? Yes, I think there is. Much of what you encounter about it, though, is irrelevant gossip. The essential is this: as has already happened in the ALM (application lifecycle management) market, vendors are moving in the direction of products that simultaneously are better out-of-the-box, and interoperate, to some extent, with plugins, extensions, or add-ons. Take advantage of this. You should only license APM products that demonstrate objective results within the first week of installation, yet also have enough extensibility to cover the customizations you’ll eventually need. Such products are available, or will be soon in 2013, depending on your detailed requirements.