The last thing you need in times of economic gloom is a cloud environment that costs more money than it should. That's why now's a great moment for businesses to step back and assess their cloud spending, then determine how they can steel it against the recession that may be on the horizon.
Keep reading for tips on preparing your cloud environment for a possible recession — and, beyond that, keeping cloud costs under control even after boom times return.
Common Cloud Spending Mistakes
The first step toward recession-proofing your cloud is to understand how companies commonly waste money in the cloud. The top threats to cost-effective cloud computing include:
- Oversizing resources: Oversized cloud resources are VM instances, Kubernetes nodes, or other types of pre-provisioned resources that are configured to use more resources than they actually need in order to host a given workload.
- Lack of cost optimization: Not taking advantage of cost optimization strategies, such as choosing discounted VM instance types, is a great way to waste money in the cloud.
- Untracked resources: Resources that you fail to track are a spending risk because you may accidentally leave them running when they're no longer needed.
- Redundant resources: Sometimes, you may deploy more copies of the same type of cloud resource than you need. For instance, perhaps you've configured multiple cloud regions when one will do, or you have redundant copies of the same workload running in different clouds.
- Using the wrong type of cloud service: Cloud computing platforms offer many types of services. Depending on your workload needs, some may be more cost-effective than others. For example, a workload that you currently host in a VM could cost less as a serverless function, if demand for it fluctuates heavily.
Best Practices for Controlling Cloud Costs
Those are the spending risks that businesses typically face in their cloud environments. Now, let's look at steps you can take starting today to get ahead of unnecessary cloud spending.
1. Examine your cloud computing bill
If you haven't looked carefully at your cloud computing bill lately, do it now. In particular, identify which types of cloud services cost you most in a given month. Then, consider whether there are alternative types of services that might lead to lower costs if you moved some workloads to them.
For instance, if you are spending a lot of money on a database service, there is a chance that you could reduce costs by moving your data to an object storage service, like AWS S3, assuming that not all of your data needs to reside in a database.
2. Find idle resources
Another basic step toward controlling cloud costs is to find unused resources, then shut them down.
You could try to do this manually by sorting through lists of running resources and determining whether they are necessary. But fortunately, most cloud providers offer tools to automate the process. For instance, you can use CloudWatch to find idle EC2 instances in the Amazon cloud.
3. Develop cloud deployment policies
Finding idle resources that are currently active is only half the battle. You should also prevent new resources from being created without proper identification and tracking.
Achieving this goal starts with defining requirements for your team members to must follow when deploying new resources. But you can also take advantage of tools such as AWS tagging policies, which can help enforce resource identification requirements for your organization automatically. When tagging happens automatically, it becomes easier to keep track of resources and avoid paying to run workloads that you don't need.
4. Scale back your cloud disaster recovery strategy
If money were no object, every business would embrace what's known as a multi-site/active site disaster recovery strategy. Under this approach, you run redundant instances of the same workload in multiple locations, so that your workloads remain available if one instance fails.
But that approach is costly. It essentially doubles your hosting costs because it requires you to host two complete instances of your workloads.
For that reason, businesses looking to cut costs may want to consider scaling back to a lower-cost disaster recovery strategy. AWS offers a good overview of more cost-effective approaches to DR.
5. Put your data in cold storage
Keeping data in "hot" storage — where it is instantly available, but where storage costs are higher — is another luxury that some businesses may decide to forgo in times of economic uncertainty. They may choose instead to migrate to lower-cost storage tiers, like AWS Glacier. The caveat is that if you want to access data stored in them, it takes some time (typically, no longer than several hours) for the data to "thaw."
Cold storage isn't the right fit for every type of data. But moving more data to cold storage tiers can reduce overall costs.
6. Consider an alternative cloud
For certain types of workloads, "alternative" clouds — meaning those other than Google Cloud, AWS and Microsoft Azure — may deliver a better tradeoff between cost and performance. That's especially true if you find an alternative cloud provider that specializes in offering just one type of service — such as data storage or VM instances. Often, these alternative providers offer better pricing than the major public clouds for the same type of workload.
Don't Delay Recession-Proofing Your Cloud
There's no simple formula for recession-proofing your cloud. But there are a variety of steps that you can take — and tools you can use — to stay on top of unnecessary spending within cloud environments. And as we sit on the precipice of a potential recession, there has never been a better time to put cloud cost optimization strategies into practice.
About the author
Christopher Tozzi is a technology analyst with subject matter expertise in cloud computing, application development, open source software, virtualization, containers and more. He also lectures at a major university in the Albany, New York, area. His book, “For Fun and Profit: A History of the Free and Open Source Software Revolution,” was published by MIT Press.
