AWS Cost Optimization (FinOps)

Cost Optimization - Financial Solution:

In cloud cost is very important factor, to manage the cost in the cloud is very crucial and challenging task. In order to manage, plan, implement, and track the cost optimization tasks as an end-to-end financial management solution. It is put into use with native AWS services. By reducing the number of instances, setting up a baseline financial model, and using a charge back mechanism, the solution allows business stakeholders to save a significant amount of money.

To migrate and transform out of support On-Premise applications to Cloud Solutions. This enablement timely decommissioning of On-Premises workloads and save associated costs.

Opportunity:

. End-to-end financial management reporting solution management.
· Cost Optimization of Cloud Application Lifecycle.
· Charge back mechanism.
· Automate Start/Stop of EC2 instances as well as RDS DB instances at
different schedules using AWS Lambda Functions, Tags which can be
saved huge cost.
· Current usage and scope of cost optimization.
· Risk associated with Cost optimization on Cloud Application with financial trend & Lifecycle tracking.

Solution:

. To find out the top spend drivers and reasons of high costs in
infrastructure.
. The cost trend analysis and various cost optimization tools like AWS Native Services.

This helps to AWS Compute Optimizer, Trusted Advisor, AWS Cost Management, AWS Budget services providing user-friendly dashboard and the tool comes with dynamic data reporting.

· To understand the current usage and scope of optimization.
· EC2 instances right sizing and baselined the instances for all kind of workloads.
· Addition of governance layer in provisioning process to reduce the increasing cost by determining the workload and defining the instance type for it.
· Terminate un-used resources based on the tracking and trend analysis.
· Automate Start/Stop of EC2 instances as well RDS DB instances at different schedules using AWS EC2 Instance Scheduler, AWS Lambda Functions, which can be saved huge cost.
· Create various life cycle rules on S3 buckets to reduce the storage cost by archiving the old storage to deletion over the period of time.
. Change the S3 bucket Standard tiering class to S3 Intelligent tiering object storage class.
· Baseline the financial model and charge back mechanism for various workloads.
· Monthly proposal for cost optimization is created using AWS Native Cloud Services.

Implementation Steps:

· With this reporting feature, we can understand the current usage and scope of optimization.
· The instances right sizing and baseline instances.
· Add a governance layer in provisioning process to reduce the increasing cost by determining the workload and defining the instance type for it.
· Terminate un-used resources based on the tracking and trend analysis. This exercise can be repeated byevery week.
· Huge cost can be saved by scaling down the instances, which are under-utilized.
· RDS DB Instances can be turned off during weekends & non-business workings hours.
· Redshift cluster can be right sized with minimum nodes for the same performance. Due to this right sizing, can be optimized a huge cost per month.
· Monthly proposal for cost optimization can be created by using AWS native services.
· Create various life cycle rules on S3 buckets to reduce the storage cost by archiving the old objects for the automatic deletion over the period of time such as after 14 days after objects uploaded in the S3 bucket.
· Baseline the financial model and charge back mechanism for various workloads.
· AWS backup policies can be implemented with less retention period.
. AWS Cloudwatch logs can be retained with minimum retention policy instead of never expiry.
. Delete the manual snapshots.
· Terminate/Release AWS resources such as EFS, ALB, NLB, VPC and associated components, OpenSearch, Elasticache, RDS DB Instances, ECS with Fargate, Direct Connect, Directory Services, Workspaces, AWS Detective, Un-used EBS Volumes, EC2 Instances, Backup Policies, S3 buckets, Auto Scaling Groups,Route53 Hosted Zones, Record sets, Health Checks.
· Set “ZERO” capacity for Auto Scaling Groups in AWS Dev/Test accounts.
. Apply the tags for Auto Scaling Group for Weekends and Non-business working hours turn off policies for non-production AWS accounts.
. Delete SNS topics, Cloudwatch logs which are unused.
. Migrate EBS Volumes from GP2 to GP3.
. Migrate from old generation instances family to new generation instances family (eg. t2, m3, m4, c4 series to t3, t4, m6, m7, c6 series).
. Migrate applications from current instances family to Graviton Processor (GP3) instance family where can be saved up to 40% cost and performance improvement.
· Migrate applications from On-Premises to AWS Cloud by using AWS Migration Tool as AWS Application Migration (MGN) Tool.
. Refactor the workloads for applications modernization such as Containerization, Microservices architecture, Serverless architecture, Kubernetes Container Orchestration.