As you migrate to the cloud, would you go all in with a pair of nines?

If you did, you probably wouldn’t let your boss know, would you?

More companies are making plans to migrate to the cloud and going all in on it. That means, first, that they’re preparing to move their existing applications to the cloud and, second, that they’re building all of their applications in the cloud from now on (a cloud-first strategy).

Now, that’s going all in. In spades.

There is a right way to go all-in on the cloud.

The right way to do it involves a series of system administration practices supported by performance monitoring software that help you optimize the resources — memory, compute, storage, networking, data protection — consumed by each application, whether in the data center or in the cloud.

A new paper from DCIG called There is a Right Way to Go “All-in” on the Cloud takes you through the practices and tools for optimizing placement of applications, whether in the cloud or on premises. It emphasizes that you can go all in without these best practices and tools, but you run the risk of overspending needlessly.

Here are a few of the highlights and takeaways from the paper.

Practices for gathering information

  1. Identify all of your on-premises applications. You can’t know how far you’ve gone unless you know where you started. Create a baseline inventory and performance level of your on-premises applications before you move to the cloud.
  2. Quantify the resources that those applications consume. Knowing the what, where, when and how much around your applications in the data center is the key to knowing what they’ll need when they’re running on a virtual machine in the cloud.
  3. Figure out what to move to the cloud and what to keep on premises. Typical triage sends test and development applications to the cloud first, followed by low-impact production applications, followed by business- and mission-critical applications. Most of the time, legacy apps and those unsupported by the cloud vendor will stay on site.
  4. Choose the appropriate type of VM instance in the cloud. By offering dozens of different types of VM instance, cloud providers prove that one size doesn’t fit all. The right type of instance for a given application is a function of memory, compute, storage and data protection requirements.
  5. Choose your cloud provider. Once you’ve collected all that information, it’s time to shop for the right cloud provider.

Surprised at how much preparation there is to going all in? You may have been tempted to jump straight to step 5 and choose your cloud provider, but if your cloud strategy is to go all in, then you’re actually better off going through steps 1 to 4 first.

Surprised at how much preparation there is to going all in? You may have been tempted to jump straight to step 5 and choose your cloud provider, but if your cloud strategy is to go all in, then you’re actually better off going through steps 1 to 4 first.

If you’re pushing all of your IT chips into the pot, you’ll want to give yourself every chance to win. That’s where continued performance monitoring comes in.

In the cloud, the meter is always running. Optimization tools help you keep a close eye on your workloads so that you never spend more than you really need to spend on keeping your applications up and your users happy. The right tool does at least three things well:

  1. Monitors how each application uses cloud resources. The tool should track compute, memory, storage, network and data protection requirements in the cloud.
  2. Provides insight on optimizing the use of those resources. Since VMs rarely need to use all their resources all the time, the tool should highlight peaks and valleys of resource usage. When usage of a given VM is low, system administrators can shut it down or move the applications from it to a more suitable type of VM instance.
  3. Works on premises and in the cloud. The tool should span both environments so that administrators see applications in the data center as they see applications in the cloud. More important, it should help measure and create the baseline performance levels mentioned above. Admins should be able to use it to make decisions about migrating apps to the cloud in the first place, or from the cloud back to the data center if need be.

Next steps

What’s your time horizon for cloud adoption? When are you going all in? Whether you’re already all in or still moving gradually toward that point, have a look at the DCIG paper called There is a Right Way to Go “All-in” on the Cloud. You’ll find more details on the points I’ve outlined above.

And if you go all in with a pair of nines and win, you’re probably lucky enough to draw to an inside straight. Just don’t let your boss know you’re doing it with your IT infrastructure.

I’m Ready to Go All In

Photo credit: Paf - Games Sport Casino

Anonymous
Related Content