One of the prime motivators for most businesses making the move to cloud is the chance to save money. It's true that you can always spin up another virtual machine (VM) and extend capacity on demand, but waiting until you need that capacity means accepting the time it takes to get the new VM online and potential processing delays. You also need to beware that the self-service capability doesn't lead to out-of-control spending when getting a new machine doesn't require the information technology team to be involved.
In fact, the key to getting the financial benefits cloud promises is to choose the right size instance for your applications in the first place and to put in place processes that limit self service.
Choosing the Right Instance Type
Every cloud provider offers varied instance types, giving choices of CPU, memory, and storage capacity. You need to start by selecting the option that suits your applications best.
This means starting by analyzing your applications and understanding their average utilization over a period of one to three months. This will let you determine the capacity they need to support their normal load and the capacity needed to support their peak load; you can then make an informed judgment about how much capacity to provide. You may also discover that some applications need to be redesigned to make use of cloud capabilities to scale efficiently.
Once you know how much capacity you need, you can start evaluating the available instances. With most cloud providers, there are configurations designed to support general-purpose computing, compute-intensive applications like analytics, memory-intensive applications such as in-memory databases, graphics-intensive applications that require GPUs, and storage-optimized instances to support transactional databases.
Because using cloud resources for short periods of time isn't expensive, you should test your proposed configuration before cutting over production usage. You'll want to simulate the workload your evaluation was based on and assure that the instance you've selected delivers the performance you need. A cloud monitoring tool can help you assess how the instance is performing.
Monitor Your Cloud Environment to Control Costs
You should continue monitoring your cloud environment even after transitioning, particularly if you've transitioned development and test environments as well as production. Dev and test are often left running when they're no longer needed, which adds to your bill unnecessarily. Forgetting to shutdown a production application that's no longer needed also results in unnecessary charges.
Don't forget to manage your people, too. Limit the number of people authorized to approve new cloud instances. While you may not want the same layers of approval needed to purchase and install on-premises equipment, you need some level of oversight to ensure that new cloud environments have adequate business justification.
Plan How to Grow Capacity
While on-demand cloud capacity lets you handle unexpected growth, if you can plan ahead you may pay less for additional capacity when it's needed. In addition to on demand capacity, Amazon Web Services (AWS) offers spot instances, reserved instances, and dedicated hosts at discounted rates; other providers have similar offerings. However, these instance types may require upfront payment or may not guarantee the capacity you need; you need to assess the trade-offs to determine the path that's best for you.
Another way to ensure you receive the financial benefits of your cloud environment is to use managed AWS services from dcVAST. Our team of experts ensures your environment is right-sized and runs optimally to deliver all the benefits of cloud. Contact us to learn more about how our services can help your cloud grow along with your needs.