Clearing Up the Cloud: Five Cloud Computing Myths Uncovered

Simply defined, the cloud is a set of compute that you attach via a network used to deliver applications and services. The cloud can take more than one form and, with that, has more interesting problems to solve in some cases. The cloud abstracts the compute and storage from the end user and delivers services (Software [SaaS], Platform [PaaS] or Infrastructure [IaaS]) that can be consumed.

While this sounds quite clear, there are still some misconceptions about the cloud. Here are some common cloud computing myths:

Myth #1: “The cloud makes everything work better.”

This is both true and false. There are many factors to think about with regard to what works better in the cloud. The cloud, like any other technology, does some things well and does others poorly. As an example, high performance databases (such as Oracle, MS SQL or Mongo) or CRM software do not run as well in the cloud in most instances. Why is this? As discussed earlier, an instance of the cloud is dependent on how it was designed; in fact, a cloud is designed for High Availability (HA), Performance, Access, Security, Sensor input and so on, or for several of them, but typically not all of the functions are built on a single cloud. The company/person designing the cloud must also answer the question, “Is this a private cloud, public cloud or a bit of both (hybrid cloud)?” Overall, there are so many factors that go into building the different cloud compute and delivery platforms that it is almost impossible to design a cloud that can run all applications, in all instances, with all interfaces to be secure, available, accountable and easy to use.

Myth #2: “The cloud provides no data security.”

Security is only as good as it was designed. Most of the data breaches in the world have been attributed to “human error.” So, in a way, this myth is both true and false. Here’s why: it really has nothing to do with the cloud, but the way the security was built into the cloud platform.

Myth #3: “The cloud is always cheaper.”

This concept is dependent on where you are standing. Do you have to build a datacenter? Do you have the staff to maintain the hardware? Do you have the DevOps team to deploy applications?

Let’s examine this more. If you look at the total cost of running in the cloud, the answers vary. As an example, if you are using a common cloud provider, some of the things you should be asking are:

Do I need a DevOps team?
What am I being charged for?
Will it perform to my needs?
Are there any time constraints to deploy my product on the provider’s cloud?
Can the provider solve my deployment issues (HA, Access, Security, etc.)?

One can also flip the coin regarding the toll of conducting business in the cloud. Charges can add up when setting up the hardware, having to install everything in a datacenter and, not to mention, getting the entire cloud infrastructure up and running to deploy your product. Finally, not all clouds are created equally. Hidden costs may sneak up on you without you noticing until it’s too late and you are locked into a vendor.

Myth #4: “Migrating to the cloud is complex and difficult.”

This is mostly false. Migrating to a cloud infrastructure enables many things including compatibility (software) and consistency, access to a wider range of services (from other companies as well as technologies) and access to your product from a wider base (employee and Customer).

There will always be a short-term pain point from a migration perspective. In the end, that small amount of pain is insignificant with the right provider that offloads and enhances what you are trying to accomplish.

Myth #5: “Public clouds are good for big data.”

Public clouds are great for what they were designed for (i.e., outsourced, “bursty” apps and cold data), but for big data — that’s a different story. The pay as-you-go model for public clouds is not the ideal solution for big data workloads. Big data clusters always need to be “on.” A stop-and-go approach wouldn’t work for this type of clusters.

Overall, there are many more cloud myths and misconceptions we could explore, but the main focus should be, “How can you take your big data workloads to the cloud?” Fortunately, Kodiak Data can help because our cloud is custom-built for big data.