Would You Plan Your Failures?
19 Jan 2010
Last week’s sales kickoff in Orlando, combined with our (Progress’) acquisition of Savvion, and my delivery of the keynote solution sales pitch (broadcast to the entire company) has left my head spinning. I’ve got a few posts in mind, but mostly have a real interest in blogging more.
I figure, best way to start is to start, and let the details work themselves out.
I realized during one of my sessions, that I’m getting kinda predictable. Every time I discuss Actional with the field, I harp on the same issue…
Automatic visibility.
Yeah, it’s the “automatic” part of that phrase that’s the important part. Other companies that provide some level of Business Transaction Management talk about visibility, but can’t tell you when something happens that you weren’t expecting (in terms of which applications are using which services). [For another view of BTM, check out OpTier’s perspective…]
I asked the audience how many people would plan in such a way that their systems failed? Not surprisingly, no one raised their hands. Who wants a failure to occur? Yet they do. And, they do because things that we didn’t know about our environment happen.
Therefore, it becomes important, and differentiating, for a Business Transaction Management solution to be able to tell us things that we didn’t otherwise know. That is, it’s important for those solutions to be automatic in how they discover what’s going on in the environment.
Let me give you a real example…
This sort of thing happened to me in several situations, but I always think of UMC in Taiwan in the year that Actional was the merged company of Actional and WestBridge (roughly 2005) when I tell the story. I installed Actional in the customers pre-production environment, and we started to display network overview unique to Actional that shows the inter-relationships between systems dynamically drawn based on actual message traffic between services. All of a sudden, and machine popped up in the top right corner of the screen. The customer, in his broken English, asked “what’s that?”
Well, it so happened that the hostname was “biztalk…”, so I knew it was a BizTalk server, and I identified the caller, the process being called in BizTalk, the calling application, etc, all the while showing him the UI and how he’d find that information using the product.
My turn to ask a question, I asked “why are you asking about that system?”
His response: “Well, that’s our production BizTalk Server on our manufacturing floor. I need to go.”
He needed to go and head off what could potentially be a major production disruption with a big business impact. When he returned about 15 minutes later, we used Actional to confirm that he had successfully shut the production system out from his pre-production systems and that there was no longer any risk that some pre-production test might take down his manufacturing facility.
You see, there are all sorts of rules/policies for making sure this stuff doesn’t happen… but with a tool like Actional you can be sure that those rules are working properly.
Before you blame this on an uncommon “process” failure, and think this was a rare event that I was lucky to capture…
Friday, we had the technical field together for a day of training, and I got to ask another question. This time, I had people raise their hands if they’ve installed Actional at a customer. Best guess, about 80 people raised their hands. Keeping their hands up, I asked…
“How many of you found stuff in your customers environment that they didn’t know was happening?”
Only two people put their hands down.
We know our customers need visibility, but that’s an incredibly powerful testament to the need for it to be automatic.
PS — I have an old presentation with 3 case studies on visibility. Email me if you’re interested, and I’ll send it over.
