An operations team manages workloads in several AWS accounts that all belong to the same organisation. They want a single CloudWatch dashboard that displays graphs built from metrics in every account, without copying the metric data into one place. Which approach lets them build such a dashboard with the least ongoing effort?
- APublish every member account metric into the monitoring account by repeatedly calling the PutMetricData API on a schedule from each source account.
- BBuild one dashboard per account in each account separately, then export the JSON and paste the widgets together into a shared document for the team.
- CEnable CloudWatch cross-account observability by linking the source accounts to a central monitoring account, then build the dashboard there from the shared metrics. Correct
- DGrant the team an IAM role in each account and ask them to switch roles in the console to view every account dashboard one at a time.
Why A is wrong: Copying metrics with PutMetricData duplicates data, adds custom scheduling to maintain, and is exactly the data movement the team wants to avoid rather than a native sharing mechanism.
Why B is wrong: Pasting exported widget JSON into a document does not create a live CloudWatch dashboard, so the graphs would not refresh and cross-account metrics would never appear together.
Why C is correct: Cross-account observability links source accounts to a monitoring account so its dashboards can graph metrics from all linked accounts directly, meeting the requirement without copying any data.
Why D is wrong: Role switching lets people view each account in turn but never renders metrics from multiple accounts on one dashboard, so the single combined view is still missing.