Are You Being Courted by a Buggy-Whip Vendor?

There was a time when buggy-whips were the height of technology. If you wanted your horse and cart to go faster, you invested in a better buggy-whip. And then came the automobile, and all of a sudden buggy-whips were obsolete.

The funny thing though, is that it wasn't instant. The companies who made beautiful buggy-whips carried on making them for quite some time and were stunned that no one was buying them. Buggy whip advertising actually grew, as companies tried harder to convince the consumer to buy more.

And in some ways, the classic application performance monitoring products are finding the same issue.

The classic model of APM relies on monitoring specific technical aspects of computing, such as memory utilization, processor consumption and latency across a network. Using these parameters allows an operator to see if things are approaching a threshold where service level agreements (SLA's) could be impacted. And it works. But it only works in some specific use cases, and things can become very complex where the interaction between many different systems has to be considered.

The classic APM trick is to advise the customer to write libraries of complex scripts to identify actually what applications are doing. This creates an expensive and risky headache for users, who must maintain this massive library of scripts through every update on the infrastructure, platforms, and applications. It's a never-ending complexity headache.

But there is a way for companies to create a level of system monitoring that goes far ahead of the classic APM model, and that is to start to make use of the messages that their middleware solutions are already sending between every component in their application stack. By intelligently using these messages, a business can now monitor exactly what is happening, and even go further and link message records together to see entire technical and even business transactions.

By looking at messages and transactions along with the classic detail of system performance, a business can now do three critical things dramatically better.

1. Monitor what is happening now (alerts based on business impacts, not just technology including Message Tracking and Transaction Tracking).
2. Understand in detail what happened in the past (to quickly identify the root cause of any issue)
3. Predict what will happen in the future (using Artificial Intelligence based on machine learning to identify likely issues well before the happen and identify likely solutions that can stop events ever happening).

These three aspects of operational control, will reduce MTTR (Mean Time to Repair), and increase MTBF (Mean Time Between Failures) and MTTF (Mean Time To Failure) in ways that were previously impossible.

So why aren't classic monitoring and APM solutions providing this level of technology. Well, they make buggy-whips, and surely people still want buggy-whips...

So how can you spot monitoring solutions that are trying to sell you buggy-whips. It's quite easy, companies that invest huge amounts in marketing and sales to sell you things that you are no longer interested in buying, who are focused on being sold to private equity or their competitors.