Bringing everyone down
Messy architecture: Who's causing your crashes?
It seems like everything is getting more complicated these days - and IT architectures are no different. Whatever you're dealing with, when things get more complex and knotty, it's much more likely to result in snaggles down the line.
A recent US study by tech and market research company Forrester has revealed that almost one third of IT organisations are running into daily issues. That’s a lot of downtime - and can cost businesses anything from $10,000 to $1 million per hour.
Ostensibly looking at the "State of IT Monitoring," the study focused on the tactics operations teams use to solve downtime, what is most important to them in a monitoring tool, and where they currently stand on addressing the complexities of the modern data center.
Even as organisations drift towards virtualised or public / private cloud environments, many of the operational aspects remain firmly tethered to localised systems such as mainframes and client servers, resulting in fragmented infrastructures. This isn’t necessarily an issue in itself, but, as Forrester’s Chris Smith points out, a fragmented approach to managing these infrastructures can bring about all manner of chaos.
According to Smith, if you take a siloed approach to management, it’s simply not possible to determine the context and correlate the root cause of issues. As he says, “it’s like looking at a single piece of evidence instead of examining the entire crime scene – you can’t solve the crime if you don’t understand the big picture.”
He told JAXenter that, “The more heterogeneous the infrastructure, the more important it is to have a holistic approach to management,” noting that an ideal approach would be to take a unified approach to managing performance and availability, regardless of the type of infrastructure in place.
Another interesting point the survey brought to light was the fact that less than half of the 127 companies are failing to recognise the strategic nature of their monitoring data - or even if they do, utterly failing to leverage it. More often than not though, organisations tend to view it as “a low-end function that exists to support incident management, i.e. getting alerts when an issue is detected.
Part of this failure to make the most of monitoring tools could be attributed to a widespread distrust amongst organisations of these technologies. Smith states that, “A key value of a monitoring tool is to generate an alert before the end users notice a problem.” However, only 62% of organizations in the study were getting this benefit. 38% had already experienced an issue before their IT team was aware there were gremlins in the machine.
Unfortunately for companies impacted by costly downtime build up, Forrester expects crashes are only going to become even more frequent as datacentres continue to evolve.
Smith says that, under the new paradigm of “hybrid IT”, the relationships “between IT assets and the services leveraging them are highly dynamic.” Without well thought out and tightly executed infrastructure management, when issues do arise, it’s extremely difficult to quickly track them down and identify culprit. And without intervention, it’s company funds that will continue to suffer.
Image by Sarabbit