Rise of the robots – can APM save us?
After my e-retailing blog last week, I have been keeping my eyes, ears and mind open to the subject of automation. Infuriated by electronic kiosks as I often am, I feel the rise of the robots may just feature as something of a theme in the months ahead.
The new ‘Terminator Salvation’ movie trailer leads with John Connor panting something like, “Oh my god! We’ve made too many machines…” – and my own experiences down the local supermarket on the auto-till when I am in the US are just as bad (OK, well, almost). I have taken to selecting “Spanish” before I pay for my milk and cookies to see if it makes the process more fun. It does not.
'
Free image use: Wikimedia Commons
Anyway, research terms like automation and auto-payments and you’ll come up with any number of vendors who claim to offer a painless and efficient route to a more ‘managed’ reality for these technologies.
For the sake of blogument, I found Precise Software to be of some interest on this topic. The company’s Executive VP Zohar Gilad has some pretty vocal opinions on application development and testing in the automation zone.
Gilad points out that while kiosks proliferate alongside ecommerce and payments made through mobile devices; more machines means more transactions – and that’s something that should strike a chord with developers and DBAs alike. CIOs will become increasingly brow beaten as the performance of machines in this zone directly impacts how well the business performs. The business-technology function divide had better close up fast suggests Gilad.
If you hadn’t guessed already, Precise sells Application Performance Management (APM) stuff (sorry, solutions). But then so does IBM, CA, Quest and a bunch of others – so what are the most pertinent issues to be aware of in this sector would you say?
I’d like to suggest that one of central truisms here is that an increasing number of enterprise applications are becoming web-enabled; this has to inexorably lead us down a path towards higher transaction volumes. So the key word vendors in this space love is VISIBILITY (alright some of them say TRANSPARENCY, but it’s the same thing in this context). This of course is to allow us to identify, understand and resolve problems quickly.
The main area of concern today appears to be focused on tools that are built to help with the analysis of what we want to know about when a transaction breaks or slow down. We need to know: what broke, where did it break, why did it break, who does it impact, how do we fix it and how do we avoid future problems of this kind?
According to Precise, “Traditional approaches to APM focus on technology silo monitoring and alerting for infrastructure problems, with no linkage to business impact. Problems still arise. Business is still impacted. The application performance problem is still unsolved. Customers experience applications one transaction at a time. Ensuring the performance of each and every transaction requires a fresh perspective in application performance management that moves beyond the silo-based approaches of the past.”
Looking ahead – it’s for sure that prevention is better than cure in this area. The more progressive APM vendors have, it appears to me, also taken the time to talk about their view on future transactions. Perhaps those made tomorrow, or perhaps those made at peak identifiable periods such as Christmas. I’m talking about ‘aggregation warehousing’ techniques to identify intermittent or seasonal trends to prevent transaction performance problems before they ever impact the end user.
Great stuff then! I’ll be happier at the auto-kiosk and the machines are not about to take over the Earth then? Don’t count on it.