Overview

Welcome to our latest blog post focusing on practical cost optimization strategies for cloud services. In this second installment, we delve deeper into the hands-on application of the pillars and design principles we previously discussed for optimizing costs.

We are excited to share a success story that showcases our expertise in action for one of our valued clients. Discover how our tailored approach helped in achieving up to 25% reduction in their costs.

Recommendations for Different AWS Services

The client’s primary objectives centered around reducing AWS cloud spending, optimizing infrastructure and applications, and establishing robust governance and best practices for sustainable cost management.

Tech Vedika conducted a comprehensive analysis of multiple client accounts generating high AWS monthly costs and provided impactful recommendations, building upon the methods outlined in our previous blog. In the following sections, we outline strategies, findings, and recommendations for typical AWS services.

1. RDS
   • We observed that most of the AWS accounts in the organization have retained manual backup snapshots unused for a long duration. Deleting the unused snapshots and limiting them to the last four days of backup resulted in a decent cost reduction.
   • Often, storage is provisioned at a level exceeding actual usage. We streamlined the storage and configured autoscaling.
   • The accumulation of data over the years in the database contributes to inefficient resource consumption. Archiving old data and retaining only a year’s worth of data resulted in good savings.
   • Leveraging Reserved instances based on usage patterns significantly contributed to cost savings.
   • For Aurora instances, transitioning from Standard to Provisioned IOPS yielded cost benefits while maintaining the performance.
2. CloudFront
   • One often overlooked setting in CloudFront pertains to VIP for SSL. Activation of this setting should only be considered if genuinely required otherwise it will result in unnecessary cost.
3. EBS Storage
   • Migrating from GP2 to GP3 volumes enhances performance and reduces costs by 20%.
   • Removing unused instances and outdated snapshots presents a substantial avenue for cost savings.
4. EC2
   • Scaling down several EC2 instances exhibiting low CPU utilization.
   • Transitioning to newer generations of instance types and incorporating Graviton instances is recommended.
5. NAT Gateway
   • Costs were significantly curbed by eliminating dormant NAT Gateways.
6. Fargate
   • Downsizing task configurations resulted in a decent reduction in cost.
7. Open Search
   • Right-sizing overly provisioned service
   • Adopting 1-year reserved instances led to big savings.
8. DynamoDB
   • Reducing RCU/WCU to match actual requirements
   • Optimizing provisioning
   • Removing redundant tables
9. Elastic Cache
   • For instances exhibiting negligible hits we suggest downsizing or changing instance types.
10. ELB
    • Eliminating unused ELBs and fixing problematic ones.
11. EIP
    • Periodically reviewing and removing unused and unassociated EIPs is advised.
12. Data Transfer
    • Shifting applications to the same region to mitigate the high costs associated with data transfer across different regions
13. Kafka
    • Adjusting disk space as it was over-provisioned.
    • Configuring autoscaling led to savings.
14. ElastiCache
    • High costs from on-demand nodes prompted:
      1. Downsizing instances
      2. Switching to reserved instances

Governance

We recommended setting up a governance team to oversee periodic cost optimization activities. The following are key activities for governance:

• Setting up a Governance team
• Implementation of cost-related monitoring, alerts, and notifications
• Establishing controls based on the budgets
• Conducting periodic reviews
• Setting up approval processes

Conclusion

Our journey into cloud services cost optimization has been enlightening and rewarding. By applying our pillars and design principles for cost optimization, we discovered various methods to enhance efficiency and achieve substantial savings in cloud operations.

As technology evolves, so do the challenges and opportunities in the cloud landscape. We firmly believe that a proactive, adaptive approach is essential to stay ahead. Embrace innovation, capitalize on data-driven decisions, and continuously refine cloud operations to unlock the full potential of your investments.

We hope that this series has provided valuable insights and actionable knowledge on cost optimization. We are committed to helping our clients in navigating the complexities of the cloud while maximizing resource value and efficiency.

For more insights, refer to our previous blog on Cost Optimization: Pillars & Design Principles.