Geek Speak

2 Posts authored by: explorevm

The reports of my death are greatly exaggerated - The On Premises Data Center Mark Twain


As I mentioned in my previous post, the hype cycle has the on-premises data center dead to rights. Public cloud is the future and there’s no turning back. When you step back and investigate the private cloud space, you’d be surprised.


When the order comes down to move to the public cloud, sometimes you discover the application or workload isn’t suitable for migration. How many times have you been in an environment where an important line-of-business application was written by an employee who left 15 years ago and has no documentation. Is the application even running on an operating system one of the public cloud providers supports? While proper planning and refactoring can provide a path to move your application to the cloud, occasionally you’ll run into unavoidable physical limitations. In manufacturing environments, often you’ll find the need for a custom interface for a machine or a USB license dongle for an application to function.


Sometimes applications can’t move to the cloud despite planning. But not everyone takes the time to plan their cloud migrations. Failure to plan leads to many issues with cloud adoption (which I’ll discuss in my next post). What about your connectivity to the internet and cloud? Whether it’s uptime or bandwidth, the network can prove to be a cloud roadblock. When these migrations fail, where does the workload go? Back to the on-premises data center!


Looking at purchasing new data center hardware or moving to the cloud often includes decision-makers outside of the IT department. While we will leave the deep fiscal details for the financial experts, Chief Financial Officers and business leaders often weigh the benefits of the Operating Expenditure versus Capital Expenditures, commonly referred to as OpEx vs. CapEx. While the ability to quickly scale your cloud resources up and down based on demand might be a blessing to an application administrator, the variations on cost accompanying it can prove to be difficult for the accounting department. The ability to make a one-time purchase every three to five years and amortize the cost over those years is a tried-and-true financial tactic.


Speaking of finances, the fiscal performance of hardware vendors must certainly be a bellwether of cloud adoption. Sales of data center hardware has to be falling with the ever-growing adoption of public cloud, right? Wrong. As has been announced over the past year, vendors such as Dell Technologies, Pure Storage, and Lenovo report record earnings and growth. In May 2019, Dell Technologies announced 2% growth year over year. May also brought a revenue announcement from Lenovo of 12.5% growth year over year. August of 2019 saw Pure Storage announce a whopping 28% year over year revenue growth. These companies are just a small example. Clearly physical data center hardware is still in high demand.


Looking at many factors, it’s easy to say the on-premises data center is far from dead. Continuing the “Battle of the Clouds” series, we’ll dive into why “just put it in the cloud” isn’t always the best solution. 



We’ve all heard it before.


“The cloud is the future!”

“We need to move to the cloud!”

“The on-premises data center is dead.”


If you believe the analysts and marketing departments, public cloud is the greatest thing to happen to the data center since virtualization. But, is it true? Could public cloud be the savior of the IT department? While many features to the public cloud make it an attractive infrastructure replacement, failure to adequately plan for its use can prove to be a costly mistake.


Moving past the marketing, the cloud is simply “someone else’s computer.” Yes, it’s more complicated than that, but when you boil it down to the basics, it’s a data center maintained by a third-party with proprietary software on top to provide an easy-to-use dashboard for provisioning and monitoring. When you move to the cloud, you’re still running an application on a server. Many of the same problems you have with your application running on-premises can persist in the cloud.


In a public cloud environment, the added complexity of multi-tenancy on the underlying resources can complicate things. Now you have to think about regulatory compliance? And after all, public cloud is still a data center subject to human error. This has been made evident over and over, famously by the Amazon Web Services S3 outage of February 2017.* The wide adoption of public clouds such as AWS and Microsoft Azure has also opened the door to more instances of shadow IT. Rogue devs, admins, and end users who either don’t have the patience to wait or have been denied resources opening cloud accounts with their own credit cards and putting corporate data at risk. And, we have yet to even take into consideration the consumption-based billing model.


Even with the above listed “issues” (I put quotes around issues as some of the problems can be encountered in the private cloud or worked around), public cloud can be an awesome tool in the IT administrator’s toolbox. Properly architected cloud-based applications can alleviate performance issues and can be developed with robust redundancies to avoid downtime. The ability to quickly scale compute up and down based on demand provides the business amazing agility not before seen in the standard data center procurement cycle. And, the growing world of SaaS products provides an easy gateway to enter the cloud (yes, I’m going to take the stance that as-a-Service qualifies as cloud). The introduction of cloud technologies has also opened a world of new application deployment models such as microservices and serverless computing. These amazing ways of looking at infrastructure weren’t possible until recently.


Is there hype around public cloud? For sure! Is some of it warranted? Absolutely! Is it the be-all and end-all technology of the future? Not so fast. In the upcoming series of posts I’m calling “Battle of the Clouds,” we’ll look at public cloud versus private cloud, going past the hype to dive into the state of on-premises data centers, what it takes for a successful cloud implementation, and workload planning around both solutions. I look forward to hearing your opinions on this topic as well!


*Summary of the Amazon S3 Service Disruption in the Northern Virginia (US-EAST-1) Region

Filter Blog

By date: By tag:

SolarWinds uses cookies on its websites to make your online experience easier and better. By using our website, you consent to our use of cookies. For more information on cookies, see our cookie policy.