A Web seminar discussion on Thursday, entitled “Infrastructure Management: New Innovations, New Pains,” featured insight from Glenn O’Donnell, an IT operations and automation analyst with Forrester Research, and Jim Rapoza, a senior research analyst focusing on application performance at Aberdeen Group. The analysts were joined by a tech officer at CA Technologies in the talk moderated by Mark Thiele of the industry community outlet Data Center Pulse.
Increasingly, applications are gaining importance over hardware, especially with the growth in cloud performance options and the business demands for more data analysis. That is forcing different layers within IT to own up to the “sins of the past” – siloed systems, isolated performance fixes and network construction that focuses on just on speed and bandwidth, says O’Donnell.
“We have been locally optimizing and we have to do some global optimization,” O’Donnell says. “Real systems engineering means we’re going to optimize the entire system and sometimes that means that parts have to be just good enough. That’s a hard pill to swallow for IT.”
O’Donnell says IT also should work together to find a primary infrastructure management offering with the awareness that you’ll also likely need a few other boutique tools to fill the gaps. That’s another move away from the past, where a few enterprises opted for “one big integrated vendor solution” that left users unfulfilled while most favored cheaper yet efficient models that missed holistic capabilities.
In demand and with increasing prowess are infrastructure management offerings dialed up in the cloud. Rapoza says the cloud can become “black holes” for on-premise network monitoring and performance tools that have often been segmented between departments, and IT has been forced to take a consolidated approach on the way that layers of data systems truly work for end users.
“There has been a scramble in the last 12 months to update those infrastructures and to still do performance management tasks,” he says.
Rapoza noted that some of the disruption from infrastructure management in the cloud comes from expectations of the agile development world, where applications are updated and tweaked more often. While slow service or poor applications in-house would be relegated to an IT department problem, applications in the cloud have the added pressure of reactions by users on Twitter or lousy reviews in an app store.
Anecdotally, Thiele noted that recent discussions he’s had with traditional big iron and hardware vendors have centered on applications and off-site monitoring, a sea change from years of on-premise investment and, in part, a response to the ultimate service that is infrastructure management. The moves toward the cloud and as-a-service piggyback on findings Gartner Research reported last week that put IaaS spending in this year at $62 billion worldwide, and growing over the next five years to the point where it nearly matches SaaS.
For a rebroadcast of the seminar, click here.