Stratusphere provides a centralized view of your organization's cloud spend and savings opportunities. The savings opportunities are presented from several data sources, including AWS Trusted Advisor, EC2 Compute Optimizer, and custom findings we've built into Stratusphere.
In this document, we will explore the custom findings rules that are supported in Stratusphere. We'll also take a look at the rule definitions, so that you understand how these items are detected in the Stratusphere Findings view.
The following custom findings rules are supported for customers with data sources connected to their Amazon Web Services (AWS) environments.
Description: The Network Load Balancer (NLB) is one of the types of load balancers available in the Amazon Elastic Load Balancing (ELB) service. This detection rule is designed to identify any ELB resources that have been provisioned, but aren't actively being utilized by any clients and services.
Criteria: To identify an idle NLB resource, we examine the Amazon CloudWatch metrics ProcessedBytes and NewFlowCount for each NLB in your AWS environment. If both of these metrics have a zero sum, across the last 7-day look-back period, then the NLB resource will be flagged as idle.
Look-back Period: 7 days
Description: The Application Load Balancer (ALB) is a layer 7 HTTP load balancer, with TLS termination support, that's part of the Amazon Elastic Load Balancing (ELB) service. This detection rule is designed to identify any ALB resources that have been provisioned in your AWS environment, but are not actively being utilized by any clients and services.
Criteria: To identify an idle ALB resource, we examine the Amazon CloudWatch metric RequestCount, for each ALB in your AWS environment. If this metric has a zero sum, over the last 7-day look-back period, then the ALB resource will be flagged as idle.
Look-back Period: 7 days
Description: Amazon Elastic Block Store (EBS) is a managed service providing block storage devices, that can be attached to things like EC2 instances and AWS Fargate tasks. Snapshots of EBS volumes can be taken at any point. These snapshots can accumulate over time and increase costs unnecessarily. This detection rule is designed to identify EBS snapshots that may be good candidates for cleanup.
Criteria: To identify an EBS snapshot as "orphaned," we look at snapshots in the standard storage tier, that were created more than 30 days ago, and is not managed by AWS Data Lifecycle Manager (DLM) or AWS Backup services. Also, the snapshot is not associated to any existing EBS volumes or EC2 Amazon Machine Images (AMI).
Description: The Amazon ElastiCache service provides several different in-memory key-value storage services. These services are commonly used for caching query responses from databases, storing user session data, tracking real-time analytics, and other similar shared storage needs across fleets of application servers. One of the supported storage services in ElastiCache is Redis Cache. This custom detection rule aims to identify ElastiCache Redis instances that have been provisioned, but aren't actively servicing any requests.
Criteria: To identify an idle Amazon ElastiCache Redis cluster, we look at Amazon CloudWatch metrics CacheHits, CacheMisses, and NewConnections, over the last 7 days. If the sum values for these three metrics are all zero, then the cluster is flagged as idle.
Look-back Period: 7 days
Description: Amazon Elastic Block Store (EBS) supports several different types of storage, depending on your performance and cost requirements. AWS released the new General Purpose SSD (GP3) storage tier, which provides similar performance to the GP2 storage tier, for approximately 20% less cost. This detection rule helps you identify EBS volumes, across your entire AWS environment, which are good candidates to upgrade to GP3.
Criteria: Any Amazon EBS volumes that are currently using GP2 storage tier, and exist in an AWS region where GP3 is lower cost than GP2, and the volume IOPS are below 16,000, are flagged under this finding type.
The following custom findings rules are supported for Microsoft Azure cloud environments that have been onboarded to Stratusphere.