Cloud Economics: How to determine your TRUE Cloud TCO- Part 3
In this post, we will discuss how to lower your cloud costs through various discounting and architectural techniques.
#1- Pre-Purchase Capacity: By leveraging pre-purchase options for compute instances, you can achieve a significant reduction in your Cloud TCO. For example, Amazon Web Services (AWS) offers an option called Reserved Instances (RIs) that can reduce your spend by up to 75 % if you use them in the same availability zone as what’s purchased. AWS offers 3 models for reserved instances:
- Standard RIs: Useful for workloads such as dev/test environments where you expect a certain standard consistent configuration for your environment. With Standard RIs you can modify Availability Zone, scope, network platform, and instance size (within the same instance type) of your Reserved Instance.Standard RIs can be used for production workloads as well, however given the restrictions with modifications, you will have limited options should you need to adjust your instances to handle new patterns of customer traffic. Standard RIs are available for 1 or 3 year terms.
- Convertible RIs: Useful for steady state workloads such as production environments. With convertible RIs you can exchange an RI for another new instance with new attributes such as instance family, instance type, platform, scope, and tenancy. Convertible RIs are generally available only for 3 year terms.
- Scheduled RIs: These RIs are typically available for a specific duration such as a few hours during the day etc. These instances are useful for running batch workloads that can be scheduled at regular intervals daily.
In addition, there are discounts available based on the percentage of upfront payment. For organizations just starting out on the cloud journey, consider reserving 90 % of your dev/test capacity using Standard RIs and atleast 80% of your production workloads on the cloud roadmap for the next 3 years using Convertible RIs.
#2-Sign Enterprise agreements:
Many cloud providers offer a significant discount for signing an enterprise agreement that includes a commit for 1-3 years. This commit is typically a $ amount that is agreed upon in terms of anticipated cloud spend. While this option is attractive, it can be challenging to determine the expected usage across an enterprise especially if you are just starting out on your cloud journey. For those who do not have a current cloud spend to base this off, we recommend going for an enterprise commit bucket that is atleast 80 % of the anticipated cloud costs for applications on the cloud adoption roadmap for the coming year.
#3- Leverage containers:
Containers allow you to do more with less. Containers operate at the Host OS level rather than at the Guest OS level (like Virtual Machines) and due to this they are able to offer better performance compared to VMs. Containers typically require a lower resource footprint compared to VMs because they share the kernel of the host with other containers allowing you to deploy multiple containers on a single cloud instance resulting in lower costs.
#4- Leverage Spot Instances: Spot instances are an AWS offering that is essentially unused capacity that AWS offers at massive discounts of up to 90 %. The only drawback of spot instances is that they can be interrupted when the capacity is no longer available and this introduces additional cloud management overhead. Spot Instances can be used to run any workload that is stateless including mission critical workloads but is often really well suited for big data analytics workloads that require massive compute power but can easily resume where they left off. Leveraging Reserved capacity along with Spot instances offers the maximum benefit and you should consider configuring autoscaling groups with spot instances for handling burst capacity.
#5- Leverage Cloud Provider OS: By leveraging OS provided by the cloud provider such as Amazon Linux , you can significantly reduce your cloud costs. Amazon linux is provided at no charge along with EC2 Instances.
Click HERE For Part 1, determining your compute, storage and networking costs.
Click HERE for Part 2, determining your add-on costs for Cloud (backup, storage, security etc).
In part 4, we will account for additional costs such as software licensing, cloud management and software licensing costs that must be accounted for.
In part 5, we will discuss determining cloud-costs in a multi-cloud environment.
(Note: All pricing examples are based on Amazon Web Services (AWS) pricing as of the date of publishing this post and are based on the US-East-Northern Virginia region)
Need Help with determining your cloud-costs and developing a business case? SIGN UP for a FREE 1-Hour Consult with us today.