ESG continues to research the interest in, advantages, and disadvantages of Integrated Computing Platforms (ICPs). As a quick review, ICP is the collapsing of server, storage, and networking infrastructure into a simple-to-consume, preconfigured, turn-key platform. We have published numerous reports on the topic including Virtual Computing Infrastructures: The Movement Toward Integrated Computing: The Foundations for Cloud and Virtualized Computing Infrastructure Preferences.
I have also been speaking about ICPs during a series of events across the country and received a fantastic question: “What do I do if it breaks?” This question came from an IT pro that, like many, operates in a do-it-yourself (DIY) environment and he felt that if anything broke, he knew how it was cabled, racked, designed, etc. He felt that this gave him an advantage to troubleshoot and fix the problem should one arise.
This was a great question and let’s all agree that ICPs will, as this attendee voiced, “break.” There are two ends of the spectrum to explore when answering this question, but before I jump into them, it’s important to understand that one of the top drivers for ICPs is to deliver improved availability, business continuity, and application uptime. By design, the goal of ICPs should be able to predict trouble and minimally be able to proactively alert IT operations. If the vendor you are working with can’t validate this, keep looking and you will finds vendors that can do this very successfully.
On one end of the spectrum, if an ICP breaks, some vendors offer the opportunity to securely connect the ICP back into the IT vendor’s service and support organization to offer proactive monitoring, alerts, and even a recovery services measured in minutes. Not all applications require this level of support, but be aware that it is available and can significantly help when things go bad.
It is also important for IT organizations to recognize that most of the ICPs available are certified, designed, and delivered as a known system to the vendor or set of vendors. This therefore should lead to expedited route cause analysis and a quick fix when an IT pro contacts support. I will say, however, that this has IT pros a little nervous and concerned if a vendor support organization says “we are giving this our best effort” and leaves IT with one vendor to work with rather than the two or three they may play against each other. Now that certainly isn’t the most efficient use of time, but is feedback I have heard multiple times.
I want to throw the question out to the ICP vendor community. What should be the expectations of an IT organization when an ICP breaks?