Banner for the cloud cost efficiency checklist blog

This is a two-part blog series which covers the fundamental questions businesses need to ask for cost-efficient cloud usage. These questions include:

  • Which is the right cloud provider for us?

  • Are our resources right-sized?

  • Are unused resources cleaned up regularly?

  • How effectively are we leveraging savings plans?

  • Are we making use of spot instances?

  • How efficient is our auto-scaling configuration?

  • Are our applications cloud-efficient?

  • How optimized are our network pathways?

While you can find the first four questions answered in the first part of this series, you can also get access to the full list by downloading our latest white paper, How IT leaders can drive more with less: An enterprise guide to technology adoption and cloud usage in a disrupted economy.

Now let’s explore the second half of the checklist:

5. Are we making use of spot instances?

A spot instance is a type of compute instance offered by cloud providers at a discounted price when there’s excess capacity available. This is cloud providers’ strategy to commercialize their unused capacity. They offer it sporadically and often take it back without much notice.

Organizations can leverage that extra power whenever it is available to accelerate business operations in a cost-efficient way. This is suitable for fault-tolerant workloads, like batch processing and media file rendering.

6. How efficient is our auto-scaling configuration? 

Auto-scaling scales out and scales in the number of instances as per configured rules and policies, based on metrics like CPU and memory.

Avoid flapping: While the scale-out mechanism adds resources to meet performance needs, the scale-in mechanism removes extra resources and saves costs. However, a faulty configuration may lead to increased costs.

E.g., the rule is to scale out one instance if CPU usage is >60% and scale in one instance if it’s <60%. If the workload frequently fluctuates around 60%, it will cause rapid scale-in and scale-out actions, called flapping. This leads to inefficient resource utilization and increased costs. To avoid flapping, it is recommended to maintain an adequate margin between scaling thresholds, e.g., add one instance when CPU usage is >60% and remove one instance when it’s <25%.

In this way, configuring auto-scaling rules is also a considerable aspect of cloud efficiency.

A cheaper way of ensuring redundancy: Maintaining multiple instances in different availability zones is necessary for high availability and fault tolerance but not for non-critical workloads, like development and testing environments, in-house usage of tools, batch processing, and internal demonstration purposes.

Here, instead of setting metric-based rules to add or remove instances, you can set the minimum, maximum, and desired number of instances to one. So, when the instance goes down, another one is created in its place. This is a cheaper way of ensuring redundancy for the non-critical workloads listed above, especially when the cost of redundancy outweighs the benefits of the workload you run.

7. Are our applications cloud-efficient? 

Rather than thinking you pay for what you use, an efficiency-driven way to think about the cloud is that you pay for what you provision. But this isn’t the case with serverless environments. Serverless environments operate at the level of individual functions, allowing for fine-grained scaling. Each function is independently scaled based on demand, so you are billed for the actual execution time of each function. This presents an excellent opportunity to save costs by rewriting applications for efficient cloud usage.  

Imagine an e-commerce platform that frequently displays product details along with customer reviews. In a traditional setup, retrieving this info might involve multiple operations across tables storing product details, customer data, and reviews. To enhance efficiency, you could consolidate the storage of product details and associated reviews into a single table. This simplifies queries and speeds up the retrieval process. The application can then quickly fetch and display product information, saving costs by minimizing database processing time.  

8. How optimized are our network pathways? 

Prioritizing essential communication pathways and blocking unnecessary traffic contributes to efficient use of network resources. By restricting access based on IP addresses and ports, you route critical data through high-bandwidth paths, while non-critical data uses cost-efficient paths.

Think about a manufacturing industry application that connects machines, sensors, and analytics services to optimize the manufacturing process. It involves both real-time and historical analysis. Here, time-sensitive (critical) data can take a direct, high-bandwidth path, while non-critical data used for historical analysis can take a cost-efficient path. You can implement this through security groups and network access control lists that differentiate between critical and non-critical data flows. 

For organizations that have evolved to a mature phase of cloud cost optimization, these efficiency-centric practices are not as much a checklist but rather part of their subconscious culture. For cloud users who are still figuring out how to tame their cloud bills, however, asking these questions can be a good start.

This cloud efficiency checklist is just a glimpse into our latest white paper, How IT leaders can drive more with less: An enterprise guide to technology adoption and cloud usage in a disrupted economy.

If your limited IT budget constrains you from realizing bigger digital transformation goals, our white paper can help you tackle it in four ways:

1. Make effective use of the budget for standard IT requirements.

Learn how our technology adoption framework can help you allocate your IT investments strategically to obtain amplified business outcomes.

2. Conventional IT investments aren’t enough. You need a growth engine, too.

Read our case study on General Electric to learn how you can apply technology on top of existing business resources and build an alternate revenue engine.

3. When it comes to cost, the cloud is a double-edged sword. Make sure you use the right edge of it.

Learn the role of CIOs in the cloud cost optimization cycle and leverage our cloud cost efficiency checklist to master the fundamentals of cloud usage.

4. Enterprise organizations are saving 60% of their cloud costs by shifting their workloads to their own data centers. Find out how.

Read the cloud exit story of a SaaS enterprise and the behind-the-scenes math of its savings in detail.

Download the white paper