Opportunity Name
High CloudWatch GetMetricData Cost (Datadog Metric Fetching)
AWS Resource Type
Amazon CloudWatch (GetMetricData API)
Opportunity Description
This Finder identifies AWS accounts with unusually high CloudWatch GetMetricData API request costs that are commonly driven by third-party monitoring integrations—most notably Datadog’s AWS integration, which polls CloudWatch metrics at regular intervals.
GetMetricData requests are billed by CloudWatch, and at scale (large numbers of resources × metrics × regions), costs can accumulate into hundreds or thousands of dollars per month.
For Datadog customers, there is a simple, proven optimization: Datadog Support can change the CloudWatch polling interval from 10 minutes to 20 minutes, which reduces GetMetricData requests by 50% and therefore cuts these costs by 50%, with minimal impact for most monitoring use cases.
This is a Finder-only opportunity. CloudFix does not modify Datadog settings and cannot change polling frequency automatically.
Criteria for Identifying the Opportunity
An opportunity is identified when:
-
CloudWatch CUR line items show GetMetricData request costs (usage types indicating GetMetricData API usage)
-
Costs exceed the minimum threshold (default equivalent of $50/month, i.e. $600/year annualized)
-
The account is not excluded via
cloudfix:dont-fix-ittagging (where applicable)
Confidence level adjustment (Datadog detection):
-
High confidence (Datadog confirmed): A Datadog AWS integration IAM role is detected in the account (role name contains “Datadog”, case-insensitive, and is used for CloudWatch read access)
-
Medium confidence: High GetMetricData costs exist but Datadog integration cannot be confirmed (could be another monitoring tool)
Potential Savings (If Known)
Savings are predictable and exactly 50% for Datadog customers who reduce polling frequency from 10 minutes to 20 minutes:
-
Current annualized GetMetricData cost: $X/year
-
Estimated optimized cost: $X × 0.5
-
Estimated savings: $X × 0.5
CloudFix reports both the current annualized cost and the estimated post-change cost based on the 50% reduction.
What Happens When the Fixer Is Executed?
There is no automatic Fixer for this opportunity.
CloudFix provides a recommendation only. The optimization must be implemented by contacting Datadog Support and requesting that they change your AWS integration polling frequency from 10 minutes → 20 minutes.
Is It Possible to Roll Back Once CloudFix Implements the Recommendation?
Yes.
Rollback is a manual process: if you need to restore 10-minute granularity, you can contact Datadog Support and request the polling interval be changed back from 20 minutes → 10 minutes.
Can CloudFix Implement the Fix Automatically Once I Accept the Recommendation?
No.
This is a Finder-only recommendation. Datadog polling frequency is a Datadog-managed setting and is adjusted through Datadog Support.
Does This Fix Require Downtime?
No.
This change does not require downtime. It may slightly reduce metric granularity (20-minute intervals), but the monitoring platform remains available during and after the change.
Additional Resources
Implementation Notes for Customers (Action Steps)
-
Confirm you are using Datadog’s AWS integration.
-
Open a ticket with Datadog Support requesting polling frequency reduction to 20 minutes.
-
Monitor CloudWatch costs for 7–10 days after the change to confirm the reduction appears in billing.
Datadog Support ticket template (copy/paste):
Subject: Request to Reduce AWS CloudWatch Polling Frequency to 20 Minutes
Hello Datadog Support,
Please reduce our AWS integration CloudWatch polling interval from 10 minutes to 20 minutes to reduce GetMetricData API costs.Datadog Org ID: [your org id]
AWS Account ID: [account id]
Regions: [regions]
Requested Polling Interval: 20 minutesThanks,
[name]
Bill Gleeson
Comments