Best Practices for Compute Engine Cost Optimization
Compute Engine helps to create and run virtual machines on Google infrastructure. Compute Engine offers scale, performance, and value that allows the user to easily launch large compute clusters on Google's infrastructure easily.
Compute Engine is an Infrastructure as a Service (IaaS) from Google Cloud Platform, which provides a customizable environment where the users can build their own Virtual Machine according to the organization's requirement. It offers services with the various OS, customized machine types (RAM, VPUs, SSD, HDD, GPU) on availability
The best practices for Compute Engine Cost Optimization are :
Reserved Instances: Compute Engine offers discounts on VMs for 1 year or 3 years. The discount price applies to the total number of VCPUs and the amount of memory usage.
Auto-Scaling: It offers the best way to reduce cost. At any point in time the organization can easily scale up or scale down compute resources depending upon actual usage.
Preemptible VMs: These are self-destructive VMs. It works best for the applications that distribute the processes among multiple instances. Preemptible VM has a short lifespan. It can run a maximum of 24 hours and even it may be stopped before 24 hours. A 30-second notice is sent to the instance when a VM needs reclamation
Auto cost-optimizations: It mainly works on the instances which are time-bound or time labelled. Suppose, a developer is using the cloud in his duty hours and after the duty hours there is no use of it. So, then there will be no usage of the service. Cloud Scheduler is mainly used for this service.
Right-sizing: A performance analysis of an organization’s workloads is essential to identify the instances that will ensure the capacity that matches the usage and therefore meet the performance requirements. All the data metrics can be gathered by using the Stackdriver Monitoring. It will check all the metrics and give recommendations for resizing the instances, which will help to optimize the total cost.
This above picture clarifies that using the Stackdriver Monitoring, the organization can reduce the cost and it will collect the data from the instances for proper usage of the resources.
This picture shows the Committed usage which helps to optimize the actual cost.