Compuware - The Definitive Guide to APM

This is an web book from RealTime publishers nexus.realtimepublishers.com Currently, only half of the chapters are delivered, so here is what it is!

The Good: Emphasis on the process gap and organization maturity as the real barriers to APM success. It includes small vignettes of IT life at the start of each chapter, to highlight to IT situation and challenges. They even define the "M" of APM as Management, not Monitoring. They avoid the word "Dashboards" in favor of visualizations (cool: brings reports back to the table) and mention "application lifecycle" a few times, like "APM optimizes the application lifecycle". And, my favorite: "measuring performance is a proxy for understanding business performance".

The Bad: It purports to "take you through an entire implementation" but doesn't offer any depth. It is more of an extended whitepaper. It does cover the lifecycle of the motivation, decisions and implementation of an APM solution but as a conversation of what could be done. They only acknowledge stakeholders as SysAdmins, Developers and End-users (the guy using the browser)

The revelation for me was that they dredged up a Gartner Maturity Model from 2003 that had some interesting contrasts with the model we derived from our internal analysis of implementation failures and successes. Gartner identified management maturity as "Chaotic, Reactive, Proactive, Service and Value". Our model allowed for "Reactive, Directed, Proactive, Service-Driven and Value-Driven", with Reactive further divided as Reactive-Negotiated and Reactive-Alerting. I don't get too much access to Gartner stuff but here is how I represented Management Maturity:

This is from the first ICMM positioning around 2005. I don't recall why I decreased the size of each box, from left to right. I think I was trying to emphasize efficiency or proportion of IT organizations that might be found practicing at that level.

"Directed" is using APM metrics to influence the application lifecycle, focusing on QA practices.

"Service-driven" - everybody has this goal, we only tried to put teeth into it by associating it the definition of best practices.

I really liked this slide but it is not the emphasis we have today. We focus more on "visibility" because it provides a more "joining" context among the various tools that are available and their contributions, rather a focus on a particular technology and excluding all others. We also highlight the impact that visibility has on the existing processes in an organization and how this helps us assess their maturity and make meaningful recommendations for remediation.

Management Maturity and the processes that go along with it are the foundation of the Compuware APM message. That's not so bad. But then they fall into some unusal partitioning of APM in order to highlight their transaction monitoring technology. So I conclude that while they are saying the right things they regrettably recast APM as something that their technology delivers - and you only need to look at transactions.

Process re-engineering is actually pretty difficult and for well-established and reasonably successful, Reactive-Management organizations, this message is just way too hollow. These prospects know a lot about processes and may even know where they have gaps. What they need instead is a plan to help them evolve the organization, consistent with APM. What can they do now, and tomorrow, (and without purchasing new technology) that will get them on track to APM? What will derail their efforts? How will they know they have improved the situation? When will investment help accelerate their drive to APM?

Like Neo (Matrix) - Stuck on the subway platform

Still waiting on the proposal acceptance. Editing assistance contract, version two, is off to legal. Some business challenges and a bit of reorganization have pulled attention away from the project. Thankfully, the marketing folks tripped over an APM book project from Compuware so now they are beating the drum to help get the oars moving on this boat!

I've started outlining sections on Triage and Maturity Models. These are both pretty big topics and left out of the initial outline but it might be useful to have them ready in case the 'editing' decimates the current block of content.

Currently, I'm at 30k words or 60% of my initial target but really didn't make much progress in October. All the ducks are lined up, in terms of the process obligations. It feels like my flight home is cancelled and I'm waiting for a schedule update - no meeting at risk - just my own time!

One Writer, Dozens of Editors

Not quite back from vacation I foolishly moved my crackberry from stun, to loud, and shortly thereafter received an invite to join an unscheduled (for me) conference call with the internal publishing group. I was a couple minutes late to the call but quickly learned that there is a lump of process to follow, none of which I had been aware of earlier. Yikes!

I'm about 50% complete with the first draft, of which such status I share when asked. I never expected to hear "Stop right now! You can't start writing until your outline is approved!" She was serious! Apparently, they expect he process to be followed and there was some concern that I would be wasting time writing now, when the outline my very well be changed.

I don't see this as much of a concern. Frankly, I would prefer to cut-n-paste from too much material, rather than come up short and have to write new material from scratch. The whole purpose of this projects was to re-purpose the work I had already done, in presentation and white paper form, and bring it all together into a single, cohesive book experience. So I'm wondering what the editors will come up with, to strengthen or reduce the outline. So far, some internal reviewers of the outline have simply determined that "It looks good. Get-er done!". My simple goal is to not be the bottleneck in this part of the process - the writing. Whatever it takes to get it into book form - I don't have any control over that. But metering the pace of writing - that doesn't seem very sportsmanlike.

But it was concluded that I needed to prepare and submit a book proposal, which would include a detailed outline, resume, and a whole lot of other stuff that the marketing folks will need when this project goes to press. And it needed to be done ASAP. So I got it together, updated my resume (which I hadn't done for over six years) and pushed out the goods Monday night.

Silence.

So far.

On the contractor front, that seems to be going well. I've chatted with my so-to-be-anointed collaborator/editor/I've_done_books-before dude. We have the first draft of their services proposal. I have that into our internal team for review and comment. The 'hot potatoes' are off in other hands. And I keep chugging along.

Part of the book proposal application was a couple of questions along the due diligence theme. Has someone else covered this topic? What publications have a similar theme? What makes your book different.

I took a moment to pause and reflect. This could be a useful bit of exercise. I've always popped off a google search, from time to time, to see if anyone has done something significant for APM and never came back with much. But them I was a little bit more biased towards performance of distributed architectures and not simply Application Performance Management.

After a concentrated effort, I starting finding a few leads. I had a moment of panic - "Crap! Has someone beat me to the punch?" But on further review, a couple of good ideas but not at all the message I'm planning to drop on the Industry. So what am I finding? I'll lay that out over the next couple blog entries.