I recently had a wonderful discussion of application performance management (APM) and business transaction management (BTM) with the good folks at INETCO. We discussed the broad concepts of APM and how an end-to-end focus on business transactions is a critical success factor. I was intending to comment on the technology behind the announcement here.
A bit about INETCO
INETCO is tightly focused on following business transactions from end to end and being able to accurately report on what is happening and where in real time. They also understand that today's workloads include components such as:
- UNIX-based midrange systems
- Single-vendor operating systems running on special-purpose hardware
- Industry standard systems running Windows or Linux
- A herd of single-function appliance servers
- Workloads that are hosted on a combination of physical, virtual and cloud-based systems
- Increasingly complex distributed, multi-tier workloads that mix and match just about everything listed above
- A constellation of company-supplied, staff-supplied and customer-supplied end-points that include PCs, Macs, Tablets, "Phablets" and Smartphones
The company believes there is value in being able to monitor the end-to-end performance of transactions rather than just offering tools that dive deeply into a huge pile of operational data.
Before I could comment
Before that commentary popped to the top of my research queue, the company's PR firm reached out to me to ask for my opinions on why a treatment of the announcement didn't appear in many analyst's publications. While I certainly can't speak for anyone other than Kusnetzky Group analysts, I do have some thoughts on the topic (surprise, surprise.) Here are a few of them.
- Regardless of how good INETCO's technology or architecture are, there is much confusion in the market for performance management tools.
- The announcement-per-second rate has gone up recently making it hard for any supplier to get a moment of decision maker's time and attention. Some really good technology is not getting the attention it desires.
- Customers are deploying an ever more complex mix of multi-tier, distributed workloads that are based upon different system architectures, hosting strategies (physical, virtual and cloud), and are allowing access to their workloads from an ever-growing array of end point systems (PCs, Macs, Laptops, Tablets and Smartphones.) It is not at all clear to them how to address the "Tower of Babel" they've built inside of their data centers at the encouragement of their chosen suppliers.
- There is a large and growing number of firms offering some sort of APM or similar products. I believe that I've spoken with representatives of over 20 suppliers in the past year.
- All of them are speaking about their products and technology using the same or very similar language.
- All of them appear to be promising the sun and the stars and some analysts (and customers) believe that in the end they are only going to deliver the moon.
- Some are very large, traditional, suppliers of management tools and others are frisky startups, each of which says that it is uniquely qualified to make sense of this mess.
- Few offer a comprehensive solution to what is increasingly becoming an intractable problem.
- Terms such as "Cloud," "Big Data," and "Predictive Analytics" are bandied about as if they are leading edge capabilities that should define one offering from another. Unfortunately, everyone is using these terms and so it is hard to separate one supplier from another.
- Often suppliers focus on only one part of the IT infrastructure and speak as if they are offering a comprehensive solution. One has to really dig into their messages to learn where the limits are. Customers don't have the time for this type of thing anymore and they can't make a mistake in selecting tools.
- Each supplier is offering a different mix of technologies and it is not clear where each fits and why.
- Building management into everything — each supplier, of course, wants their own tools and frameworks built in and don't always play well with those offered by others.
- Install agents everywhere that gather operational data and forward it elsewhere for processing and analysis
- Follow network traffic and tease out operational data for systems, databases, application frameworks, applications, networks and storage
- Being able to audit and secure complex solutions is an ever-large factor in tool selection
What other factors do you think could be involved?