Still trying to figure out the difference between IaaS and PaaS? We can help. Watch the webinar replay here.
Ask three colleagues what cloud computing is, and you’ll get three different answers. Ask three cloud providers, you’ll get three different pitches. Three consultants, three different strategies… All these different explanations could be correct. The cloud isn’t a one-size-fits-all solution, so there are few absolutes to grasp if you’re looking for some solid ground to stand on as you do your research.
A Quick History of Cloud Computing
The history of cloud computing sheds some light on the confusion surrounding cloud. Let’s jump back in time for a moment to around 2010. Virtualization of servers via VMware had reached critical mass in thousands of IT organizations. IT and the business finally reaped the benefit of getting more value out of less server hardware and the ability to manage applications from a distance. But there were still the unpredictable problems of managing, troubleshooting, and repairing the underlying hardware. To better serve the business, IT wanted more.
Just a few years earlier in 2007, AWS had introduced what is now referred to as the public cloud: multi-tenant but secure compute resources were available at the swipe of a credit card and users didn’t have to manage the underlying hardware anymore. Many early adopters, especially developers, left the frustrations of working with the infrastructure team behind.
But public cloud was new, and many business stakeholders and C-suites didn’t trust it. Data breaches and outages were big news in the early days, and skeptics used them as cautionary tales to stay away from the cloud. Many IT people still liked the on-premises data center and the control it offered. They were often worried that cloud computing could eliminate their job.
Now let’s head to 2013 so we can witness the introduction of the private cloud. IT teams improved their virtualization of servers with newer software that allowed for a more cloudlike experience for their applications and end-users. But the lingering problem of maintaining the underlying server and network hardware still limited the value by not significantly reducing IT’s workload. At the same time, a few service providers also started to deliver VMware-based cloud infrastructure.
As all this was happening, another form of cloud computing was emerging: software-as-a-service, also known as SaaS. Modern, born-in-the-cloud applications became available to businesses without the headache of owning and operating server and data center hardware. SaaS is common now and is the delivery method of choice for collaboration, HR, and CRM applications like Office 365, Workday, and Salesforce.
Still, many IT organizations found that even though they could leverage new cloud services and applications, they needed some systems to remain on-premises. Some legacy applications and latency-sensitive applications aren’t a good fit for cloud services and maintaining them on-premises is still a better option. This model, often called hybrid cloud, is extremely common today.
Some organizations were able to eliminate the local data center by moving applications to a combination of SaaS, hyperscale cloud, and private cloud. This form of hybrid cloud has come to be known as multi-cloud.
Public cloud, private cloud, hybrid cloud, multi-cloud. It’s a lot to manage. And the industry is constantly shifting. At times it can feel like a conspiracy theory designed to make you keep buying more services.
So How Do You Make Sense of All This?
Don’t overreact and don’t overthink. Think sensibly, and make sure you work to eliminate complexity wherever possible.
First, let’s begin with a simple definition of cloud: A shared set of resources, managed by someone else, on a consumption pay model.
Next, be aware that most organizations will have applications in multiple clouds. But how these multiple clouds are used is the key to success or failure. Let’s define what we believe to be the most sensible approach to using multiple clouds: multi-cloud is using different clouds for their best purpose to achieve optimal business outcomes.
When you’re looking to minimize risk and maximize resiliency, the thought of running the same applications on multiple clouds can seem alluring. However, this version of multi-cloud holds the same risks as trying to maintain two homes and use exactly 50% of each. Expenses are astronomical and the added complexity to make it work outweighs the assumed resiliency you would get, so this isn’t exactly the optimal way to approach multi-cloud.
The better approach is this: You must assess your application portfolio one by one. Analysis of your applications allows you to evaluate unique characteristics like sensitivity to latency, the volume of network traffic/chattiness, real-time performance requirements, application interoperability, and server resource usage. Then choose the best cloud for each application. Cloud experts can use these characteristics to help you map a strategy to use multiple clouds without cost overruns and unnecessary outages.
By utilizing and optimizing multiple clouds to manage performance, costs, and to minimize non-essential effort for your IT team, your team can reap the long-term benefits of cloud.
Now, say one of your colleagues asks you what cloud computing is. Your answer will probably be different than the other three you asked before. But that’s ok. All these different explanations could be correct. The cloud isn’t a one-size-fits-all solution.
Watch the webinar replay here: