Independent insights · SAP SuccessFactors · The future of work

Amazon DevOps Guru Closes to New Customers Today: The 2027 Shutdown Migration Checklist

Amazon DevOps Guru Closes to New Customers Today: The 2027 Shutdown Migration Checklist

Effective-date alert: 29 October 2026 | Original AWS notice: 29 September 2026 | End of support: 30 September 2027

An AI-powered monitoring service is entering its final year. Amazon DevOps Guru stops accepting new customers on 29 October 2026. Existing customers can continue using it during the transition, but AWS says the service will stop operating on 30 September 2027. This is not a rumor or a same-day product launch: AWS's transition documentation dates the original announcement to 29 September 2026, and today's milestone is the new-customer cutoff.

For teams running business-critical workloads on AWS, including SAP-connected applications, the most important risk is not that servers suddenly stop. They will not. The risk is losing anomaly alerts, incident insights, API integrations, and the ability to update infrastructure-as-code stacks that still reference DevOps Guru resources. AWS sets out those consequences in its official end-of-support notice.

Four dates to put on the operations calendar

29 September 2026: AWS's documented announcement date. 29 October 2026: DevOps Guru closes to new customers. 30 June 2027: AWS's recommended target to complete migration and validate replacements. 30 September 2027: final end of support. AWS lists the complete timeline in its transition guide.

Existing enrolled accounts retain service access during the transition, including security patches, critical fixes, and AWS Support. AWS says it will not add new features. The eligibility of any particular account should be confirmed directly in the AWS environment.

What breaks after the shutdown?

Running EC2 instances, databases, applications, and SAP workloads are not shut down by this change. DevOps Guru is an observability layer. However, after the cutoff, its console and APIs will be unavailable, previously generated insights cannot be retrieved, and DevOps Guru-specific resource types will be withdrawn. The consequences for a monitoring integration or deployment pipeline can still be severe.

For infrastructure-as-code teams, AWS specifically identifies AWS::DevOpsGuru::ResourceCollection, AWS::DevOpsGuru::NotificationChannel, and AWS::DevOpsGuru::LogAnomalyDetectionIntegration. A withdrawn resource type can cause an entire CloudFormation stack update to fail. Equivalent Terraform resources, CDK definitions, and CLI scripts also need inventory. This is why AWS ranks infrastructure-as-code remediation above dashboard cleanup.

CloudWatch plus DevOps Agent is not a one-click replacement

AWS recommends Amazon CloudWatch for metric and log anomaly detection and alerting, with Amazon DevOps Agent for AI-assisted investigation. The split matters: DevOps Guru automatically analyzed resources in a coverage boundary, while CloudWatch requires teams to choose metrics, create anomaly detectors, set alarm behavior, and route notifications. DevOps Agent investigations are started from an alarm or on demand rather than replicating every DevOps Guru behavior automatically.

AWS's feature-by-feature comparison also calls out a gap: DevOps Guru's proactive insights do not have a direct one-to-one replacement. Teams should recreate critical early warnings with utilization and service-quota alarms, then test whether the resulting alerts actually arrive and trigger the intended response.

What about DevOps Guru for Amazon RDS?

The database-specific functionality retires with the rest of the service. AWS points customers to CloudWatch Database Insights for database load and performance analysis, including alarms on DBLoad and relevant wait-event metrics. That is particularly relevant to teams operating Amazon RDS and Aurora estates alongside ERP integrations, custom applications, and data services. It does not mean SAP HANA databases are automatically covered by this RDS-specific replacement path.

Review the existing DevOps Guru for RDS documentation and the AWS database transition guidance before changing monitoring ownership.

A practical migration plan for platform and AMS teams

  1. Inventory every account and Region. Identify where DevOps Guru is enabled, which applications it monitors, the alert recipients, and the teams that depend on its insights. Record any SAP-facing integrations without assuming DevOps Guru is part of the SAP product itself.
  2. Search repositories and deployment templates. Find DevOps Guru CloudFormation resource types, Terraform definitions, CDK constructs, CLI commands, and API calls such as ListInsights and DescribeInsight. Assign owners to each dependency.
  3. Rebuild detection in CloudWatch. Select important application and infrastructure metrics, configure anomaly detection, create alarms, and connect SNS or EventBridge notifications. For logs, enable CloudWatch Logs anomaly detection where appropriate.
  4. Test investigation workflows. Evaluate Amazon DevOps Agent with representative incidents, access policies, telemetry, and approval controls. Do not assume every legacy recommendation maps to a new agent action.
  5. Export historical evidence. AWS advises using SNS to capture new insights and APIs to export past insights before they become unavailable. Retain evidence according to internal audit and retention requirements.
  6. Run side by side, then retire deliberately. Validate new alert coverage against live events, remove DevOps Guru declarations from infrastructure-as-code templates, and only then disable analysis coverage and retire channels. Aim to finish by 30 June 2027, not the final shutdown day.

What I would tell an SAP customer

Do not open a blanket SAP migration project because of this announcement. First establish whether your AWS operations team actually uses DevOps Guru to monitor infrastructure supporting SAP integrations or other business-critical workloads. If it does, this is an observability and automation migration, not an SAP application migration. The acceptance test is simple: can your team still detect, investigate, and route the same critical incident after DevOps Guru is removed?

My priority would be deployment-pipeline dependencies first, alert continuity second, and historical insight export before final offboarding. The official AWS retirement notice and detailed migration checklist are the sources of truth. The shutdown date is verified; whether your organization is affected remains a customer-specific check.

No comments:

Post a Comment