Cloud Cost Optimization and Reduction Best Practices

The cloud offers organizations unlimited scalability and lower IT costs by only charging for the resources we use.

But in fact the truth about Amazon Web Services (AWS) pricing and Microsoft Azure pricing is that cloud customers pay for the resources they order, whether they use them or not.

Consequently, I have compiled a list of following best practices to optimize and reduce the cloud costs that I use in day to day operations.

• Extensively use Serverless technologies (SNS/Notification Hub, SQS/Service Bus Queues, API Gateway, EKS/AKS with Fargate/serverless etc.) as well as services such as Lambda/Azure functions instead of servers provisioning for ad hoc and small compute processing and infra admin activities that imply Pay as you go model
• Continuously monitor and delete unused resources and storage such as EBS/Azure Managed disk and Snapshots
• Automate process of shutting down of systems (EC2 instances/VMs and RDS/Azure SQL Database/Database on VM) during weekdays and completely shutting down on weekends in Non-Production environments
• Continuously monitor the resources utilization using AWS Native tools like AWS Trusted Advisor/Azure Advisor and right size the resource as needed or delete the resource if not required at all
• Continuously monitor AWS cost explorer/Azure Billing services which gives the cost breakup of each services daily and monthly which is a vital component for effective planning
• Setup lifecycle policies on storage and log retention on log groups
• Use Provisioned throughput capacity with autoscaling or On-Demand capacity for DynamoDB tables or Azure CosmosDB serverless
• Use Reserved Instances for long term commitments which may reduce the cost upto 70% cost
• Use AWS Cloud Formation/Azure Resource Manager templates that saves a significant time to provision services and environments and indirectly reduces the cost

Please do reach out to me in case of any queries or anyone needs any assistance for implementing either of these techniques.