Case Study: How Flatiron Health Gained Visibility and Control Over Total Platform Costs

Exploring AWS Cost Allocation Tags and Best Practices

In recent years, cloud computing has moved from a value-add to a must-have for many businesses looking to stay competitive with technology. As cloud adoption increases, the use of cloud platforms has grown significantly. According to Gartner, Amazon Web Services (AWS) remains the market leader in cloud platforms with more than 45 percent of the market share in 2019.

Infrastructure as a service provides businesses with a competitive advantage with its scalable and adaptive cloud computing infrastructure and storage on demand. As business requirements change, every organization needs to remain agile, and cloud computing drives agility in modern software systems.

Though these flexible cloud solutions are great for agility, they make it challenging to track resource allocation and cost. The best way to gain effective visibility into cloud resources is by assigning tags to those resources.

Let’s start with understanding tags, then we’ll look at how to apply tags to resources in AWS and how to use them to manage cloud spend and utilization.

Introduction to Tags

A tag is a label assigned to a cloud resource. Tags are composed of two parts: a key and a value.

Think of a tag like a name tag: the key, for example, could be “name”, and its value could be “James Bond”. In large organizations, operating keys may refer to different departments, like “finance” and “operations”, and their corresponding values may be more specific projects, uses, or environments like “staging” or “production”.

There are two different types of tags. The first kind of tag is provided by AWS, such as subnet ID or instance ID, for which AWS automatically generates long strings of letters and numbers. The second type of tag is user-defined, allowing users to identify resources in a way that makes sense for the business. For example, a large organization may define a particular tag’s key as “environment” and its value as “production”.

The terms tag and label are interchangeable depending on the service provider or technology in use. Tags assign metadata to AWS resources so the consumer can use the key-value pair to filter, manage, and search for the resources.

Tags enable AWS users to gain deeper insight into their resource use. With solutions like Yotascale, they can allocate those resources to the particular business units, products, and development teams to better understand which projects are driving costs.

Categorizing, Identifying, and Managing Resources

Tags help effectively categorize AWS resources based on their environment, owner, or purpose. For example, assigning different tags to Elastic Compute Cloud (EC2) instances of your Amazon account enables you to track stack-level tags as well as each instance’s owner. Similarly, these tags also help categorize different servers, for example, application servers, data servers, middleware servers, and more.

Tags can also script the AWS execution or automate functions. For example, you know to shut down any server with the code tag “nightly” after 7:00 PM. Tags help you easily and effectively manage resource use, and otherwise challenging and time-consuming task.

Calculating Cost

Tags help large organizations calculate the cost of AWS resources. Let’s say an organization has a large AWS footprint, with most departments using Amazon services, resulting in a large monthly bill. To calculate costs, simply use the cost-centered tag on all resources to identify the particular department using the resource. You can then determine that resource’s billing amount and allocate it to the correct department’s budget.

Using tags enables you to use cost allocation in AWS Cost Explorer. Easily categorize different services using tags. For example, if you want to focus more on a specific business area, tags make it more feasible by enabling you to analyze the cost of resources required for those particular business operations. However, for more complex environments such as multiple accounts or containers / Kubernetes, you’ll need to look beyond AWS Cost Explorer to help you manage costs.

Challenges of Resources Without Tags

We’ve examined the benefits of using tags as they provide a deeper insight into resource use. Let’s now analyze the challenges that may arise if we are not using tags.

For example, say your account has no resources right now. A developer comes in and starts launching resources for your business unit. After some time, you feel the need to add more resources. Developers need to share some information for these different resources so they can go ahead and share some common databases, EC2 instances, and more.

Then, over time, some old servers come into the picture that may include notifications and lambda functions. All these resources keep adding up, and each of them has a unit cost. This all causes a huge cost build-up until finally the organization gets a huge bill. This bill may generate a few questions, like:

These questions are challenging to answer if your cloud resources are not tagged. The best way to manage resources is to assign metadata to the AWS resources in the form of tags.

Example Use Case

Let’s now consider a use case where a company benefits by using tags.

An organization is working with various environments in the cloud-like staging, development, testing, and production — and there are various resources available for each environment — like compute storage and lambda functions. A problem arises at the end of the billing cycle, when the cloud bill is much higher than anticipated. The bill offers no indication of where this unexpected cost was incurred, so the accounting team is not sure which department is responsible and cannot correctly track budgets. Additionally, the organization has no way of knowing how to prevent similar cost spikes in the future.

This problem can be avoided by placing resource- or service-specific tags for resource identification. Tags can help identify or isolate the costs incurred. Properly allocating the resources to their real owner makes the owner accountable for the cloud spend.

As a bonus, these tags can help in saving even more money, by identifying extra spend on idle resources. Tags provide greater visibility into resource use and costs, enabling owners to optimize their cloud resource use.

Tagging also enables audits of confidential data and can identify all the resources requiring additional security scrutiny. This is invaluable for organizations that require payments, or store medical information, subject to PCI audits or regulated by the Health Insurance Portability and Accountability Act (HIPAA).

Wrapping Up

Cost attribution and tagging, when used correctly, can do wonders for your business. Tagging AWS resources helps you track your cloud spend and easily find the information you are looking for.

Yotascale’s cloud cost management solution can help with tagging hygiene by setting up tagging policies consisting of mandatory flags for tag keys and regular expression validation for tag key values which identify and remediate past inconsistencies. For example, tags that were named similarly but are not identical, tags with spelling errors or tag keys missing values,

Yotascale can also automatically inherit from the parent or update tag key and tag key values with auto-tagging rules that are established within the platform. This becomes crucial when you are working with dynamic environments such as Kubernetes that can scale on demand.

To learn more about Yotascale, and optimize your AWS Cloud budget, request your Yotascale demo today.