Supported automated insights
To access automated insights, you must have an Enhanced or Enterprise DoiT Cloud Intelligence plan. You may need additional configurations or plans, depending on the source of the recommendations and insights. See Feature availability for more information.
AWS insights
From cost and usage data
| Service | Insight | Description | Category |
|---|---|---|---|
| AWS CloudTrail | Optimize CloudTrail Event history costs | Use only organization trails to avoid charges for duplicate event records. | FinOps |
| Amazon Elastic Block Store (EBS) | Migrate from gp2 to gp3 | This insight estimates the potential savings you can achieve by converting your EBS gp2 volumes to gp3 volumes. | FinOps |
| Amazon Elastic Block Store (EBS) | Delete old EBS Snapshots | This insight identifies EBS snapshots in the Standard tier older than 90 days as candidates for cost optimization. | FinOps |
| Amazon Elastic Kubernetes Service (EKS) | Delete unused EKS clusters | EKS clusters incur usage charges on an hourly basis, regardless of whether they run any workloads. | FinOps |
| Amazon Elastic Kubernetes Service (EKS) | Upgrade EKS clusters | Effective from the April 2024 billing cycle (starting April 1, 2024), Amazon EKS clusters running a Kubernetes version in extended support will be charged $0.60 per cluster hour. | FinOps |
| Amazon Elastic Load Balancers | Delete idle Elastic Load Balancers | Elastic Load Balancers incur hourly usage charges when they are running, regardless of whether they receive any traffic. | FinOps |
| Amazon ElastiCache | Migrate ElastiCache nodes to Graviton | Graviton nodes provide better performance and upwards of 5% savings compared to their x86 equivalents. | FinOps |
| Amazon ElastiCache | Migrate ElastiCache Redis OSS nodes to Valkey | Amazon ElastiCache supports Valkey with Serverless priced 33% lower and node-based priced 20% lower than other supported engines. | FinOps |
| Amazon EMR | Migrate Elastic MapReduce instances to Graviton | Graviton instances can provide better performance and upwards of 15% savings compared to their x86 equivalents. | FinOps |
| Amazon Kinesis Data Streams | Delete idle Kinesis Data Streams | Kinesis Data Streams incur hourly charges, regardless of the throughput. We consider a Kinesis Data Stream idle if it hasn't received any traffic in the last month and has been provisioned for over 336 hours (14 days). | FinOps |
| Amazon Neptune | Migrate Neptune instances to Graviton | Graviton instances can provide better performance and upwards of 5% savings compared to their x86 equivalents. | FinOps |
| Amazon OpenSearch Service | Migrate OpenSearch instances to Graviton | Graviton instances can provide better performance and upwards of 10% savings compared to their x86 equivalents. | FinOps |
| Amazon RDS | Migrate RDS instances to Graviton | This insight looks into your current x86 spend and estimates your savings based on the average savings generated by migrating to Graviton. | FinOps |
| Amazon RDS | Upgrade RDS engine versions | RDS instances running on engines beyond their standard support end date incur charges for RDS Extended Support. | FinOps |
| Amazon S3 | Avoid S3 high early deletion charges | This insight identifies early deletion charges that are above 1% of your total cost of AWS S3. | FinOps |
| Amazon VPC | Avoid unallocated Elastic IPs | This insight identifies Elastic IPs that are not allocated to a Network Load Balancer, NAT gateway or an EC2 instance. | FinOps |
| Amazon VPC | Delete idle NAT gateways | NAT gateways incur hourly usage charges, regardless of whether they receive data transfer. We consider a NAT gateway idle if it hasn't received any traffic in the last month and has been running for over 336 hours (14 days). | FinOps |
| Amazon VPC | Deploy VPC gateway endpoints | Reroute AWS-bound traffic to reduce NAT Gateway Data Processing charges. | FinOps |
From AWS Trusted Advisor
DoiT platform interacts with the AWS Trusted Advisor API to perform operations on Trusted Advisor checks you're entitled to. Currently, DoiT Insights supports all the checks of the following categories:
-
Cost optimization: Insights generated from the cost optimization checks fall into the category of FinOps.
-
Security: Insights generated from the security checks fall into the category of Security.
From AWS Cost Optimization Hub
DoiT platform interacts with the AWS Cost Optimization Hub API to generate resource-level recommendations.
Insights provided by AWS Cost Optimization Hub fall into the category of FinOps.
| Resource | Insight | Description | Category |
|---|---|---|---|
| Amazon EC2 instances | Migrate EC2 instances to Graviton | Graviton instances provide an average of 20% price-to-performance uplift compared to their x86 equivalents. | FinOps |
| Amazon EC2 instances | Rightsize/upgrade EC2 instances | Moving to different instance types can offer significant cost savings, depending on your usage. | FinOps |
| Amazon EC2 instances | Stop EC2 instances | Stopping idle or unused resources will save up to 100% of the resource cost. | FinOps |
| Amazon EC2 Auto Scaling groups | Migrate EC2 Auto Scaling groups to Graviton | Graviton instances provide an average of 20% price-to-performance uplift compared to their x86 equivalents. | FinOps |
| Amazon EC2 Auto Scaling groups | Rightsize/upgrade EC2 Auto Scaling groups | Moving to different instance types can offer significant cost savings, depending on your usage. | FinOps |
| Amazon EC2 Auto Scaling groups | Scale in EC2 Auto Scaling groups | While EC2 Auto Scaling itself has no additional fees, scaling in idle or unused resources helps reduce EC2 instance costs. | FinOps |
| Amazon EBS volumes | Delete EBS volumes | Deleting idle or unused resources will save up to 100% of the resource cost. | FinOps |
| Amazon EBS volumes | Rightsize/upgrade EBS volumes | Moving to different volume types can offer significant cost savings, depending on your usage. | FinOps |
| AWS Lambda functions | Rightsize Lambda functions | Adjusting memory and vCPU settings on Lambda functions can offer significant cost savings, depending on your usage. | FinOps |
| Amazon Elastic Container Service (ECS) tasks | Rightsize ECS services | Adjusting memory and vCPU settings of Amazon ECS tasks can offer significant cost savings for long-running tasks, depending on your usage. | FinOps |
| Amazon ECS tasks | Delete ECS services | Deleting idle or unused resources will save up to 100% of the resource cost. | FinOps |
| Amazon RDS DB instances | Migrate RDS DB instances to Graviton | Graviton instances provide an average of 20% price-to-performance uplift compared to their x86 equivalents. | FinOps |
| Amazon RDS DB instances | Rightsize/upgrade RDS DB instances | Moving to smaller instance types can reduce costs for idle and over-provisioned RDS DB instances (rightsizing), while moving to later generation products (upgrading) can increase the performance of workloads. | FinOps |
| Amazon RDS DB instances | Stop RDS DB instances | Stopping idle or unused resources will save up to 100% of the resource cost. | FinOps |
| Amazon RDS DB instance storage | Rightsize/upgrade RDS DB storage | Moving to different storage classes and changing and changing provisioned IOPS settings can reduce costs over-provisioned RDS DB storage volumes (rightsizing), while moving to later generation products (upgrading) can increase the performance of workloads. | FinOps |
Google Cloud insights
From billing data and BigQuery audit logs
| Service | Insight | Description | Category |
|---|---|---|---|
| BigQuery | BigQuery Lens On Demand Savings | Switch to BigQuery editions to reduce cost of on-demand BigQuery jobs not covered by reservations. | FinOps |
| BigQuery | Cluster and partition BigQuery tables | We analyze your BigQuery jobs from the last 7 days for suboptimal table partitioning and clustering configurations. You can access the detailed insights and take action directly via the BigQuery Recommendations Console. | FinOps |
| Cloud Logging | Optimize logging costs | Logging costs above 10% of total spend could be optimized. | FinOps |
| Cloud Storage | Move from multi-region GCS bucket to single-region bucket | Single-region buckets offer a 33% savings over multi-region buckets. | FinOps |
| Cloud Storage | Avoid GCS High early deletion charges | Choose a suitable storage class to avoid high early delete charges. | FinOps |
| Cloud SQL | Upgrade Cloud SQL instances | Cloud SQL instances running a MySQL or PostgreSQL major version that reaches end-of-life (EOL) will be charged for extended support starting on May 1, 2025. | FinOps |
| Compute Engine | Move from Performance SSDs to Balanced SSDs | Balanced SSDs offer 41% savings over Performance SSDs. | FinOps |
| Google Kubernetes Engine (GKE) | Delete unused GKE clusters | GKE clusters incur usage charges on an hourly basis, regardless of whether they run any workloads. | FinOps |
| Google Kubernetes Engine (GKE) | Optimize underutilized GKE node pools | Re-evaluating the instance type or adjusting the autoscaling policy for node pools with low utilization can optimize costs. | FinOps |
| Google Kubernetes Engine (GKE) | Rotate GKE clusters' credentials before they expire | This insight requires permissions for Kubernetes insights at the organization or project level. It becomes actionable if we detect that your GKE cluster credentials will expire within the next 90 days. | Reliability |
From Google Cloud Recommender
A recommender is a service on Google Cloud that provides usage recommendations for Google Cloud resources.
| Resource | Insight | Description | Category |
|---|---|---|---|
| vCPU and memory | Purchase resource-based Committed Use Discounts | CUDs provide significant savings compared to on-demand pricing. | FinOps |
From Google Cloud Recommender BigQuery Export
| Resource | Insight | Description | Category |
|---|---|---|---|
| Container | Remove idle clusters | Idle GKE Standard clusters incur costs even when not in use. | FinOps |
| VM instances | Stop Idle VMs instances | Identify VM instances unused for the past 14 days to help avoid charges for uptime. | FinOps |
| VM instances | Resize VM instance machine types | Resize your VM's machine type based on CPU and memory usage, ensuring efficient resource allocation and cost savings. | FinOps |
| VM instances | Enable database auditing | Database auditing allows you to monitor specific user actions in the database to help with security and compliance. | Security |
| Persistent disk | Stop idle disk instances | Idle Persistent Disk volumes in a RUNNING state incur charges for uptime. | FinOps |
| External IP address | Remove idle IP addresses | Identify IP addresses unused for the past 14 days to help avoid charges for uptime. | FinOps |
Snowflake insights
| Service | Insight | Description | Category |
|---|---|---|---|
| Snowflake | Optimize Snowflake auto-suspend | Set auto-suspend to 60 seconds for your warehouses that are used for sporadic, light queries or in cases where immediate response times aren't critical, such as in development or testing environments. | FinOps |