Last week I met with a company called dynaTrace, an organisation that labels itself as practitioners in the art of Continuous Application Performance Management, or APM if you prefer. I actually wrote about this company relatively recently, so was quite keen to get under the skin of this technology to try and decide for myself whether this was panacea or puff – and whether the company’s argument for inceased application fragility holds water.
dynaTrace says that its technology is inherently dynamic and that it extends APM to diagnostics and prevention throughout the application lifecycle – in development, in test, in staging and in production. Initially I had been somewhat critical of the company’s broad-brush statements relating to the importance of collaboration, however well intentioned they might have at first appeared.
Collaboration is of course undeniably important and breaking it down so that collaborative information is available to different stakeholders at different times – as different information means different things to different people – is a sensible idea. While system architects and testing might be interested in SQL statements and message calls, business managers might simply want to know which search criteria are scoring the most hits on the company web site.
dynaTrace is aiming to nestle itself neatly inside this ‘information share’ zone with its PurePath transaction tracing technology which aims to provide a root cause analysis of application performance. dynaTrace’s Roger Andrews and Andi Grabner told me that they aim to, “Make transparent the inner workings of Java/.Net applications, visualising application dependencies and models.” This they say, along with automated issue documentation, will eliminate the need for issue reproduction.
You may or may not question the need for analysis at this granular level; my guess is that most readers generally will not. dynaTrace’s Andrews argued that as IT becomes further embedded into every business process layer, then the application itself is becoming more fragile. If it breaks, the consequences are greater. Being able to assess the impact of a problem, isolate the application component causing that problem and identify its root cause all become even more essential governing factors – but being able to do all this dynamically throughout the application lifecycle is they key.
Essentially, I suppose this is all about looking for developer inefficiencies – and it is certainly also about being able to perform this kind of analysis whether you are working with virtualised or physically deployed applications. If dynaTrace can be criticised for anything it may be over-use of acronyms and buzzwords. But trying to be slightly kind, that may be because of the technology space that they are in. This is Business Transaction Monitoring with real-time business-level to code-level visibility – see what I mean?
The company runs a hard core developer-zone style blog if you have interest in reading more. This is the type of technology that will allow developers to scrutinise applications to examine whether a particular process takes six seconds when it should take four – and this is important when you are trying to examine production processes with an eye on trying to meet and deliver to Service Level Agreements.
So anyway, nice one guys – good to meet you, interesting stuff. Next time go easy on the acronyms and try spelling your company name with a capital letter.