OPS05-BP04 Assess operational readiness
An operational readiness review (ORR) is a structured check that your team can operate a workload safely before it goes live and as it evolves. It uses a checklist built from best practices and past incidents to surface gaps while they are still cheap to fix.
Desired outcome:
-
Your organization has an ORR checklist that captures its operational best practices.
-
Your teams run an ORR before a workload launches to general availability.
-
Your teams re-run the ORR periodically across a workload's lifecycle to catch drift.
-
You feed lessons from past incidents back into the ORR checklist so failures do not recur.
Common anti-patterns:
-
A team launches a workload without confirming it can operate it.
-
Governance and security requirements are left out of launch certification.
-
Workloads are never re-evaluated after launch.
-
Workloads launch without the required operational procedures in place.
-
The same root-cause failure appears in multiple workloads.
-
Operational readiness checklist is treated as a manual document that is filled out from scratch at a single point before launch and then left stale as the workload evolves.
Benefits of establishing this best practice:
-
Workloads launch with architecture, process, and management best practices in place.
-
Lessons from service disruptions and other impacting events feed into the review process.
-
Required procedures exist before a workload launches.
-
Reviews continue across the workload lifecycle, not only at launch.
Level of risk exposed if this best practice is not established: High
Implementation guidance
An ORR has two parts: a process and a checklist. The process is a self-service review your team runs to certify its own workload against a set of requirements. Your organization must adopt the ORR and back it with an executive sponsor. Without that mandate, teams skip readiness reviews under deadline pressure. The Operational Readiness Reviews (ORR) whitepaper describes the practice in detail. Read it before you build your own.
A strong checklist covers architecture, operational process, event management, and release quality. It also includes your security, governance, and compliance requirements. The items that matter most come from your own post-incident analysis. At Amazon, the Correction of Error process is a major source of ORR items.
Run the ORR before a workload reaches general availability. Re-run it periodically to catch drift from best practices as the workload changes. Many teams keep two variants: a fuller checklist for new launches and a lighter one for periodic reviews. As your cloud practice matures, encode recurring requirements as architecture defaults and automated controls, so those checks don't need manual review.
An ORR produces discoveries. A discovery with a mitigation already in place can pass, but without a mitigation, it becomes a backlog item to resolve before go-live. Services such as AWS Config, AWS Security Hub CSPM, and AWS Control Tower controls can detect many checklist items automatically.
Where possible, connect your readiness checklist to live signals such as alarm configuration, infrastructure-as-code definitions, and deployment pipelines, so draft answers stay current without being compiled by hand at each review. This lets you catch drift while a workload is still being built. Generated answers are suggestions, not verified facts. A person must accept, decline, or correct each one, and an unreviewed answer counts as unanswered.
Implementation steps
-
Gather the key stakeholders: Bring together representatives from security, operations, and development to define what operating this class of workload safely requires. Their combined view keeps the checklist from reflecting a single team's blind spots.
-
Collect requirements from each stakeholder: Ask each stakeholder to contribute at least one requirement, and cap the first iteration at about thirty items so the review stays usable. The example ORR questions in the ORR whitepaper are a useful starting set.
-
Consolidate the requirements into a checklist: Collect the requirements into a single shared checklist. You can use custom lenses in the AWS Well-Architected Tool to build the ORR and share it across your accounts and organization.
-
Select a workload to review first: Choose one workload to pilot the ORR, ideally one that is pre-launch or internal. A lower-stakes workload lets you refine the checklist before you apply it broadly.
-
Run the checklist and triage discoveries: Work through the checklist and record each discovery. A discovery is acceptable when a mitigation is already in place. Any discovery without a mitigation goes to the backlog and is resolved before launch.
-
Iterate and expand the checklist over time: Add new requirements and best practices as they emerge, and incorporate lessons from post-incident analysis. As your practice matures, encode recurring checklist items as automated controls using services such as AWS Config, AWS Security Hub CSPM, and AWS Control Tower controls.
-
Automate checklist evidence from live signals: Connect your checklist tooling to monitoring data, alarm configuration, and infrastructure or code repositories. Pre-populate items with draft answers and linked evidence, refresh them on a regular cadence, and route each draft through a human accept or decline step before it counts as satisfied.
Resources
Related best practices:
Related documents:
Related examples:
Related services: