Cloud waste has become a major challenge for any organization. A rapid shift to the cloud, coupled with the complexity of multi-cloud management, makes it difficult for companies to control cloud spend and eliminate waste.
According to Anodot’s 2022 State of Cloud Cost Report, which surveyed 131 US-based IT directors and executives, 88% of respondents ranked optimizing and reducing existing cloud resources as extremely or very important. Almost half of the IT executives surveyed said they struggled with cloud cost management, and their organizations wasted at least 33% of their cloud budget.
The report highlights an ongoing cloud visibility crisis in which organizations find it difficult to gain a full picture of their cloud usage and costs. As a result, they cannot control spending and reduce waste — even though cloud migration is on the rise and cutting costs is a priority.
The problem is compounded by a deteriorating economy and a potential recession, which forces companies to accelerate cost cutting and increase profits.
What is cloud waste?
Cloud providers charge for services provisioned, even if they’re not used. When cloud services are not used or are underutilized, cloud waste occurs. Furthermore, On-Demand rates for resources are significantly higher than commitment-based rates, resulting in additional cloud waste.
The amount of wasted cloud spending is often underestimated by companies. The first step to tackling the cloud waste problem is to define what waste is.
Cloud waste can be roughly divided into four categories:
1. Resources that can be right-sized
Overprovisioned resources are the most common source of cloud waste. Developers often ignore rightsizing when spinning up resources for the first time.
Rightsizing is the most effective way to control cloud costs. This involves analyzing instance performance and usage patterns, and then optimizing overprovisioned or poorly matched instances by switching to a more suitable instance type which will cost less.
2. Resources that are idle or unattached
Provisioned cloud resources that are not being used are considered idle. AWS charges for provisioned resources regardless of usage — this includes services like DynamoDB, ElastiCache, Elasticsearch, Elastic Load Balancing, etc.
AWS also charges for resources attached to stopped instances such as EBS volumes, S3 objects, etc. Additionally, AWS will charge you a small hourly rate for any IP that is not associated with a running instance.
Another source of waste is orphaned volumes and snapshots. When an EC2 instance is terminated, the attached EBS volume(s) often fall through the cracks, and the provisioned storage continues to be billed. Additionally, EBS snapshots continue to incur data transfer fees and S3 storage costs until they are deleted.
3. Commitment purchase opportunity
There are multiple ways to pay for cloud resources: On-Demand, Savings Plans, Reserved Instances, Committed Use Discounts, and Spot Instances. Using On-Demand is the most expensive option.
Adopting flexible pricing models that offer discounts based on commitment or availability is a great way to reduce cloud waste. With cheaper payment options, like Spot Instances (up to 90% off On-Demand prices) or machines covered by a commitment agreement, the hourly cost significantly decreases.
4. Incorrect configuration
Many developers overlook basic configurations for a variety of reasons, which can lead to unnecessary waste.
Previous generation instances. AWS regularly introduces new types of instances according to application loads, memory, storage and networking capacity, which always offer optimized performance and pricing.
Operating systems. EC2 instances running on Windows Server are expensive. Unless there are specific requirements for another OS, Linux operating systems are cheaper.
Unnecessary data transfer. AWS charges for inter-Availability Zones data transfer between Amazon EC2 instances within the same region. To avoid unnecessary data transfer charges, co-locate instances within a single availability zone whenever possible.
Wrong EBS type. AWS offers a variety of EBS volume types, each with its own performance and pricing. Changing from an io1-8700 GiB to a gp2 can save you around 50% each year without compromising performance.
S3 multipart uploads and versioning. You pay for storing objects in your S3 buckets regardless of usage. If multipart uploading is enabled, you are charged for incomplete multipart objects until they are deleted. When versioning is enabled, S3 rates apply to every version of an object stored and transferred. Every version of an object is the whole object, not the delta from the previous version.
How to reduce cloud waste
The good news is that cloud waste can usually be prevented. It requires a combination of a cloud cost management strategy, called FinOps, and a cloud cost management platform to provide deep visibility, continuous monitoring, recommendations, and forecasting.
The FinOps framework is an evolving cloud financial management discipline and cultural practice designed to bring accountability to cloud spend and enable organizations to get maximum value by helping engineering, finance and business teams to collaborate on data-driven spending decisions.
As resources change constantly, monitoring and optimizing cloud costs is an ongoing process. Cloud cost optimization platforms like Anodot can help you understand and control your cloud spending and continuously eliminate cloud waste. They can scan for dozens of savings opportunities automatically and calculate a sustained waste percentage. As a result, you can measure whether the overall waste of all cloud costs has been reduced, or if a specific type of waste has been reduced.
An integrated and proactive approach to managing cloud costs provides a higher return on investment for every dollar spent, and long-term, cost-effective operations.
Cut cloud waste with Anodot
Anodot has developed the most advanced recommendation engine available on the market today. The engine tracks your usage data, utilization metrics, and pricing options across AWS, Azure, and GCP to support your cloud cost management strategy, and pin-point and prioritize optimization efforts.
Anodot’s recommendation engine allows our customers to take continuous action to avoid waste, overprovisioning, and save millions of dollars every day.