中文
devexamples
← Workflows

阅读中文版本

Reserved Capacity Review

Operate cloud infrastructure safely: review cost anomalies, verify failover, and confirm resources are tagged and within quota.

intermediate~30 mincloudcostreliability

Objective

Catch cost and capacity anomalies early and confirm the failover path works before an incident forces it.

Environment

  • A working cloud environment with access to the target system and its tooling
  • Configuration and credentials for the cloud workflow installed and verified
  • A rollback or undo path documented before any change is applied

Steps

1. Review cost anomalies

Compare the current spend against the forecast and investigate any service whose cost deviates beyond the threshold.

Expected state: Anomalies are explained or corrected and the forecast is back in line.

2. Verify failover

Trigger a controlled failover to the secondary region and confirm the service stays available with acceptable latency.

Expected state: Failover completes within the RTO and data loss stays within the RPO.

3. Confirm tagging and quota

List untagged resources and compare usage against quota with enough headroom for the next growth step.

Expected state: All resources are tagged and headroom is documented for each quota.

Decision points

When should failover be tested in production?

  • During a scheduled drill with no active incidents
  • Only during a real incident

Test during scheduled drills; the first failover should never happen during a real incident.

Expected outcome

Catch cost and capacity anomalies early and confirm the failover path works before an incident forces it. The change is verified and a rollback path is confirmed.

Verification

Catch cost and capacity anomalies early and confirm the failover path works before an incident forces it. Compare before and after metrics and confirm the expected state is reached.

Cleanup

Archive the run artifacts, remove temporary credentials, and update the runbook with any changes.

Overview

Operate cloud infrastructure safely: review cost anomalies, verify failover, and confirm resources are tagged and within quota. The goal is to catch cost and capacity anomalies early and confirm the failover path works before an incident forces it.

aws ce get-cost-and-usage --time-period Start=2026-10-01,End=2026-10-25 --metrics BlendedCost

Steps

  1. Review cost anomalies. Compare the current spend against the forecast and investigate any service whose cost deviates beyond the threshold. Expected: Anomalies are explained or corrected and the forecast is back in line.

  2. Verify failover. Trigger a controlled failover to the secondary region and confirm the service stays available with acceptable latency. Expected: Failover completes within the RTO and data loss stays within the RPO.

  3. Confirm tagging and quota. List untagged resources and compare usage against quota with enough headroom for the next growth step. Expected: All resources are tagged and headroom is documented for each quota.

Verification

Confirm the outcome by checking: catch cost and capacity anomalies early and confirm the failover path works before an incident forces it. Run the verification command and compare the result against the expected state. If any check fails, stop and investigate before continuing.

Decision points

  • When should failover be tested in production?
    • During a scheduled drill with no active incidents
    • Only during a real incident
    • Recommendation: Test during scheduled drills; the first failover should never happen during a real incident.