Suvish Viswanathan is the senior analyst, unified IT at ManageEngine, a division of Zoho Corp. You can reach him on LinkedIn or follow his tweets at @suvishv. This is a first part of a three-part series, with this article focusing on "The Evolution of Infrastructure Management."
Over the years, the challenges associated with data center management have grown at an exponential rate. Once we might have looked out and imagined that the challenge was primarily one of scale: How do you fit all the computing and storage assets into the fixed footprint of the data center itself? Some data center veterans will remember seemingly endless rows of VAX systems and Winchester disks stretching out like the crates in the warehouse at the end of “Raiders of the Lost Ark.” But then those big crates were flattened into pizza box-sized things, and we started growing upwards. Suddenly, there were individual cabinets with voluminous racks of servers — and, for a while, it seemed that ability to scale would remain unchecked.
But that was just one perspective. The sheer number of computing assets we could now accommodate in the physical data center, including network devices, applications, servers and storage devices, put more pressure on the day-to-day management practicalities. Once, you could slip a disc into a machine, tap a couple of keys and patch a software issue or update an entire application. Disc in, disc out, done. But what soon followed in the amazing expanding data center sent chills up the spines of IT managers: disc in, disc out, done. Repeat.
And repeat again.
Soon it was clear that we had to come up with whole new ways to manage these highly interdependent assets in the data center. Either it took too long for one person to patch all the machines or it took too many people to do it all in the [arbitrarily defined] amount of time it used to take.
Enter the Age of Managing Management
The list of "Things To Be Managed" just goes on, evolving in sync with the demand for data and computing resources — firewalls, load balancers, network devices and more. In fact, by this time, we’ve realized that the task of managing the IT assets of the data center is so complex that we need tools to help us manage the management tasks themselves. We need change management tools that can help us keep track of what has to be done, who did what, when it was done and more. We need tools that can enable us to roll back quickly to a prior state if we find something went awry. We need tools that can model the impact of changes and that can monitor the state of this complex environment so that the heartbeat of the data center remains steady and strong.
Note, too, that so far we’ve said nothing about managing non-routine maintenance and update issues. This is just day-to-day complexity. Factor in the unexpected failure of a computer or storage device and that can reprioritize every management task we’ve just mentioned.
Moving into the Virtual World
In some ways, today’s data may seem simpler than its predecessors. The sheer number of physical servers may be lower than it was a decade ago. But each of the servers remaining on the data center floor may be playing host to tens or hundreds of virtual servers, which stretches the management complexity of this environment into whole new realms. Which virtual machine (VM) is running where? Is it at capacity? Overloaded? Underutilized?
Even the nature of the data center itself has evolved. The data center may no longer belong to a single company. An independent service provider may own the data center and provide outsourced services to companies from all over the world. Who defines the policies and practices and service level agreements that apply to the assets under management? How are those policies and practices managed, at the machine level, at the VM level, within various private cloud environments that the data center may be hosting?
Finally, just to add a little more complexity to the mix (because, clearly, we can never get enough complexity), we now have to spend more and more time thinking about how the assets in a data center are performing relative to other assets and applications over which we have virtually no control. Mobile and social media can place huge and unpredictable demands on IT resources. Hyperconverged applications pose similar problems, yet become more and more the norm (assuming that the government gets the kinks out of healthcare.gov and demonstrates the viability of this modality). We’ll need tools to help us respond quickly and appropriately as demand ebbs and flows in those domains.
Rising to the Challenge
Thankfully, there are companies that have created IT management tools that are capable of providing the comprehensive support that today’s complex IT environments require. They can facilitate management of the physical, virtual and cloud-based IT infrastructures. They can manage the seemingly innumerable configuration changes in each of these environments. Some management tools even include in-depth analytical tools that can identify root causes and troubleshoot them automatically — and that can alert the right IT resource when encountering an issue that cannot be fixed programmatically.
The only problem is, that’s still not enough.
In the modern data center, IT management is only half the challenge. The other half of the challenge lies in the management of the data center infrastructure itself. All these IT assets require an enormously sophisticated physical infrastructure, with power considerations, security considerations, cabling considerations and more. Successful delivery of the IT services to the end customers who need it depends entirely on the successful management of the data center infrastructure itself. Traditionally, that’s been handled by another team entirely, using an entirely different set of tools and an entirely different set of management and performance criteria.
That’s a problem. It’s a problem we’ve been trying to overcome for some time through awkward alliances between IT management and data center management organizations. We’ve seen the equivalent of celebrity marriages formed between high-profile IT and data center management firms, each promising to work together to solve a client’s data center challenges. But those marriages last as long as most celebrity marriages last, and when the marriage breaks up, it is the client that has depended on that partnership that suffers most. Their data center is left unsupported by one or the other party.
Yet, I remain hopeful of a solution to this problem: a truly integrated approach to IT and data center infrastructure management, an approach that enables data center managers to manage both the IT assets and the data center infrastructure assets through a single, integrated tool and enables optimal delivery of services to the customers who rely on the data center.
But more on that in my next post. Stay tuned…
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.