Whether you’re an IT admin or an IT director, even if it isn’t your number one priority to move data and workloads to the cloud, then it’s probably still in your top three. At Qumulo, we hear regularly from both customers and prospects that “the cloud” weighs heavily on everyone’s mind lately.
From a logistics perspective, many IT departments are challenged by space limits in their legacy data centers, and by an increasingly dispersed workforce. From a strategic perspective, the industry is shifting more and more of the world’s data processing to the cloud, and there are a lot of reasons for individual organizations to follow suit. Some of the more common reasons: to simplify collaboration across different organizations and geographies; to take advantage of the ever-expanding suite of cloud-native applications and services, or any number of initiatives particular to an organization.
Wherever the cloud fits in both your short and long-term plans, you’re clearly not alone in assessing your options.
Thrust vs. drag, lift vs. gravity
Like a plane in flight, being pushed up at the same time as it’s pulled down, and pushed simultaneously forward and back by opposing forces, today’s IT organizations are not only grappling with strategic and operational pressures to move operations into the cloud, they’re also trying to navigate a number of opposing forces holding them back.
This is the first in a series of blog posts that will explore some of the considerations around moving your workloads to the cloud – the pressures, the obstacles, the risks, and the issues. We want to take an honest look at what today’s enterprises are trying to do, as well as what they’re trying not to do, and how difficult it can be to do the right thing for your business AND your bottom line.
So, let’s take a look at some of the reasons why enterprises might be looking to move their workloads to the cloud. Obviously, your mileage may vary, but broadly speaking, there are a number of pressure points that are impelling enterprises to look at the public cloud as a long-term solution to them.
Data center factors
Many on-prem data centers that are still in use today are 30-40 years old, and were never designed for the kind of workloads that are common in today’s enterprises. It’s been a long journey from the mainframe/midrange systems and tape libraries of the late 1980s and early 1990s to the hundreds of racks and thousands of servers that convert electricity into heat in those same legacy data centers.
There’s no more floor space for more racks, no room for more chillers, and the power circuits are already beyond capacity. And here’s the crux of the problem: data centers are stuck in time, while technology marches on. The businesses that own those data centers need new services, features, a new competitive edge – where are they going to put them?
- What are your options when one of your business units has a critical deadline looming that you can’t support with your existing on-prem infrastructure? “More servers” are not feasible when they won’t arrive in time, or when there’s no place to put them.
- While many enterprises have existing infrastructure and plans to ensure business continuity in the event of a major disaster event (or even a short-term sitewide outage), many of those leverage secondary data centers that have the same shortage of capacity for new workloads. And yet, those new workloads keep coming, and the critical ones still need a place to fail over.
- On the subject of new workloads – besides needing an external DR/BC environment, many of them also come with one or more pre-production environments for UAT/QA validation. Those environments also require servers, storage, and other infrastructure, and those have to go somewhere too.
For all the different directions that the latest technological developments are going, they’re all leading away from the on-prem, owner/operator data center models that have prevailed for the past four (count ‘em – FOUR!) decades.
Of course, data center issues are just one of the challenges faced by today’s IT teams. Another is the fact that today’s workforce is increasingly remote from a few years ago. The pandemic of the past few years only accelerated what was already a trend: more knowledge workers than ever are now either working from home full time or 3+ days of every week.
That’s one piece of the story. Here’s another: many of those remote knowledge workers are the same ones who often need reliable, low-latency access to large amounts of data as their on-prem peers. Some teams are not only fully remote, but the individual members of a single team can literally be dispersed all around the world. While it’s possible to build a high-powered virtual desktop infrastructure capable of satisfying some very high performance demands, it’s very difficult to build it in an on-prem data center.
It isn’t just business contributors either. Even IT management and staff can fit this description. A number of companies have made the decision to send their entire IT department to work from home due to a shortage of office space. In those cases, there isn’t really any value from having an on-prem data center at all anymore.
Of course, we can’t overlook the ways in which applications and workflows have evolved. For decades, there were two primary workflow types: those that leveraged block storage (e.g. databases and virtual machines), and those that were built around file storage, like shared folders, video production, and medical imaging.
Over the past few years, Amazon, Azure and GCP have all rolled out a number of proprietary, cloud-native applications that enable powerful new workflows that would have been difficult, if not impossible, for enterprises to achieve through in-house development. There are a couple of caveats, however. For one, Amazon’s cloud-native apps are only available in Amazon’s AWS service, not on-prem (and Azure apps are restricted to Azure customers in Azure tenants, etc.). For another, cloud-native apps have been architected around neither block nor file data, but are instead based on object storage.
There’s a vast ocean of possibilities in these cloud-native apps, but they’re incompatible with the type of data that most enterprises are used to working with, and they’re definitely not available in the on-prem data center.
Infrastructure on demand enables more focus on innovation
Many of the issues that today’s enterprises are facing are directly resolved by moving workloads to the cloud. The capacity limits of yesterday’s data centers are pretty much eradicated by the virtually limitless compute, network, and storage resources across AWS, Azure, and GCP. Many remote knowledge workers are already benefiting from the low latency access that a global cloud presence can be easily configured to enable, through some combination of geo-proximal caching and follow-the-sun replication.
Not only that, but cloud adoption proactively confers additional benefits that we haven’t even touched on yet. For one, there’s virtually no lead time when it comes to provisioning new compute and/or storage capacity: a few clicks for one-off provisioning; or simple scripts for repeatable or large-scale cloud deployments, and you can be up and running in literal minutes. Compare that to the weeks (or months!) it can take between submitting a purchase request for new hardware and actually taking delivery of it. Think of what you could do with the time you reclaim from not having to deploy and maintain your own infrastructure.
We’ll examine the financials in more detail in a future post, but if you’ve ever had to manage a budget for your data center investment, you already know how difficult it can be to plan and account for your capital outlay over a 36-60-month window. Cloud’s operational-cost model renders a lot of the TCO associated with IT services much more transparent.
Cloud is inevitable – or is it?
That’s definitely what many IT strategists and architects and senior management are considering. In fact, search engine volume for cloud storage has increased dramatically in just the past few months. So, it isn’t just you: there are a lot of reasons driving this surging interest in enterprise IT.
So, cloud, right?
If only it were as simple as breaking out the corporate P-card, adding a few new entries to your routing table, and then watching magic unfold. Unfortunately, nothing is ever that simple, and for every force pushing your enterprise’s metaphorical airplane toward the cloud, you still have to acknowledge and address the resistance that’s pulling you back. That’s a topic for the next blog post, though. Stay tuned!