AWS Security ChangesHomeSearch

AWS wellarchitected: ADVOPS03-BP02 rewritten as operational readiness review gate for launches

Service: wellarchitected · 2026-11-11 · Documentation low

File: wellarchitected/latest/video-streaming-advertising-lens/advops03-bp02.md

Summary

The best practice page was replaced: instead of guidance on automating runbooks (scaling, third-party disruption, capacity, cost, data management runbooks with Systems Manager), it now prescribes a formal operational readiness review gate before publisher workload launches, with checklists, a named accountable owner, load testing at peak, rollback dry-runs, and formal sign-off. Related best practices and documents links were also updated.

Security assessment

The change is purely operational-process guidance (readiness checklists, monitoring coverage, rollback dry-runs, named owners) with no reference to a specific vulnerability, exploit, credential, encryption, or access-control weakness; it is reliability/operational-risk content rather than security remediation, so it is not tied to any concrete security incident or CVE.

Evidence

+# ADVOPS03-BP02 Establish operational readiness reviews for publisher launches

Diff

diff --git a/wellarchitected/latest/video-streaming-advertising-lens/advops03-bp02.md b/wellarchitected/latest/video-streaming-advertising-lens/advops03-bp02.md
index ce9fc7e..e01fc4a 100644
--- a//wellarchitected/latest/video-streaming-advertising-lens/advops03-bp02.md
+++ b//wellarchitected/latest/video-streaming-advertising-lens/advops03-bp02.md
@@ -7 +7 @@
-Implementation guidance
+Implementation guidanceResources
@@ -9 +9 @@ Implementation guidance
-# ADVOPS03-BP02 Automate runbooks to gain operational efficiency
+# ADVOPS03-BP02 Establish operational readiness reviews for publisher launches
@@ -11 +11 @@ Implementation guidance
-Document runbooks for failover procedures, capacity scaling, and incident response workflows using AWS Systems Manager documents for automation. 
+New SSP adapter integrations, SGAI pipeline deployments, FAST channel launches, and innovative ad format rollouts must undergo an operational readiness review before entering production. The review is a formal gate. A workload that fails any checklist item must not activate. 
@@ -13 +13,14 @@ Document runbooks for failover procedures, capacity scaling, and incident respon
-## Implementation guidance
+**Desired outcome:**
+
+  * Every publisher workload launch has verified monitoring coverage, documented runbooks, defined escalation paths, and validated rollback procedures before serving live traffic. 
+
+  * The readiness review is a formal gate with a named owner accountable for sign-off. 
+
+  * Load testing has been completed at expected peak traffic levels before activation. 
+
+
+
+
+**Common anti-patterns:**
+
+  * Launching new SSP adapter integrations without verifying monitoring coverage. The most common manifestation is discovering after launch that there are no alarms for adapter timeout spikes. 
@@ -15 +28 @@ Document runbooks for failover procedures, capacity scaling, and incident respon
-Consider the following example use case runbooks: 
+  * Deploying SGAI pipelines without documented playbooks for investigation of novel failure modes, leaving on-call engineers to improvise investigation procedures during incidents. 
@@ -17 +30 @@ Consider the following example use case runbooks:
-**Scaling runbook**
+  * Skipping load testing before FAST channel launches. Live events drive traffic spikes that may exceed design capacity. 
@@ -19 +32 @@ Consider the following example use case runbooks:
-  * Design step-by-step workflows for manually scaling up and down Amazon EC2 instances and increasing or decreasing managed service capacities 
+  * Treating readiness review as a checkbox exercise rather than a genuine gate that can block activation. 
@@ -21 +33,0 @@ Consider the following example use case runbooks:
-  * Create automation scripts to initiate auto scaling actions based on predefined events 
@@ -23 +34,0 @@ Consider the following example use case runbooks:
-  * Perform validation checks for successful scaling operations 
@@ -25,0 +37,14 @@ Consider the following example use case runbooks:
+**Benefits of establishing this best practice:**
+
+  * Reduced risk of unmonitored production failures. 
+
+  * Faster incident response through pre-validated runbooks and on-call training. 
+
+  * Improved team confidence during launches. 
+
+
+
+
+**Level of risk exposed if this best practice is not established:** High 
+
+## Implementation guidance
@@ -26,0 +52 @@ Consider the following example use case runbooks:
+Publisher ad tech launches routinely proceed without formal readiness validation. Engineering completes a feature, ad operations confirms it works in staging, and it goes live. In the first production incident, no monitoring dashboard shows the relevant metrics, no runbook exists for the failure mode, and the on-call engineer is debugging a system they have never operated in production. 
@@ -28 +54 @@ Consider the following example use case runbooks:
-**Third-party service disruptions**
+A readiness gate helps you discover problems earlier. Every useful checklist item captures a lesson from a past incident. A checklist built built from your own past failures and problems is shorter and more effective, especially when it is kept up to date with your environment. 
@@ -30 +56 @@ Consider the following example use case runbooks:
-  * Implement multi-provider redundancy and failover mechanisms using AWS Lambda functions and Amazon API Gateway 
+An item belongs on the checklist only if it could plausibly fail. Questions that are answered by default can produce false sign-offs that do not certify anything about your workload. Useful checklist examples include whether the on-call engineer has operated the workload before, whether a failure mode has a documented response, and whether a recovery path has been exercised. 
@@ -32 +58 @@ Consider the following example use case runbooks:
-  * Use [AWS X-Ray](https://aws.amazon.com/xray/) for end-to-end tracing and troubleshooting of distributed applications and third-party integrations 
+Baselines are often skipped but are the most expensive item to skip, because you can't create them retroactively. After a launch, you can only judge whether current behavior is normal against a record of what preceded it. You can capture a baseline before activation for almost nothing, but you can't reconstruct it afterward. 
@@ -34 +60 @@ Consider the following example use case runbooks:
-  * Document playbooks for provider switching, data synchronization, and incident escalation using [AWS Step Functions](https://aws.amazon.com/step-functions/) and [AWS Lambda](https://aws.amazon.com/lambda/)
+A checklist becomes a control only when one person owns it. When a team owns the checklist, no specific person verifies each item, because every item looks like another member's responsibility. Naming one owner also assigns accountability when a launch fails after passing. 
@@ -35,0 +62 @@ Consider the following example use case runbooks:
+Launches fail at the peak. Testing at expected load leaves that risk untested. The margin above expected peak reflects how wrong the forecast could be. Set it generously for scheduled events where a third party controls the programming. 
@@ -36,0 +64 @@ Consider the following example use case runbooks:
+### Implementation steps
@@ -37,0 +66 @@ Consider the following example use case runbooks:
+  1. **Define a standardized publisher operational readiness checklist template:** Create a checklist template covering: monitoring dashboards configured and displaying expected metrics, runbooks documented for well-understood failure modes, playbooks documented for novel failure investigation, on-call rotation includes staff trained on this workload, load testing completed at expected peak traffic levels, rollback procedures fully tested, and a designated readiness review owner with sign-off authority. 
@@ -39 +68 @@ Consider the following example use case runbooks:
-**Infrastructure capacity issues**
+  2. **Customize the checklist per publisher workload theme:** Extend the standard checklist with workload-specific requirements. Prebid adapter additions: timeout threshold configuration, bid response validation, yield baseline metrics captured. SGAI additions: full-pipeline validation, SCTE-35 validation active, CDN cacheability verified. FAST channel additions: multi-AZ MediaLive configuration, EPG completeness validation, live event load testing. 
@@ -41 +70 @@ Consider the following example use case runbooks:
-  * Implement auto scaling and load balancing using Amazon EC2 Auto Scaling and [Elastic Load Balancing (ELB)](https://aws.amazon.com/elasticloadbalancing/)
+  3. **Assign a readiness review owner for each launch:** Name a specific individual (not a team) as the readiness review owner who is Responsible for verifying all checklist items and Accountable for the sign-off decision. This owner must have the authority to block activation if checklist items are incomplete. Document the owner's name in the launch planning record alongside the expected activation date. 
@@ -43 +72 @@ Consider the following example use case runbooks:
-  * Use [AWS Auto Scaling](https://aws.amazon.com/autoscaling/) for predictive scaling based on historical data and scheduled scaling for planned events 
+  4. **Conduct load testing at expected peak levels:** For FAST channel launches, run load tests at 150% of expected peak concurrent viewer count. For Prebid Server launches with a new SSP adapter, run synthetic auction traffic at 2x the expected adapter bid rate for 30 minutes. Validate that all monitoring dashboards accurately reflect workload behavior under load and that no unexpected alerts fire during the load test. 
@@ -45 +74 @@ Consider the following example use case runbooks:
-  * Document runbooks for capacity planning, scaling procedures, and cost optimization using [AWS Systems Manager](https://aws.amazon.com/systems-manager/) Documents 
+  5. **Execute a rollback dry-run and require formal sign-off:** Before activation, execute a full rollback dry-run: deploy the workload to production, then trigger the rollback procedure, verifying that traffic reverts to the pre-launch state within the target time. Obtain formal written sign-off from the readiness review owner only after all checklist items are verified including the rollback dry-run. Document the sign-off with timestamp in the launch record. 
@@ -50 +79 @@ Consider the following example use case runbooks:
-**Cost optimization runbook**
+## Resources
@@ -52 +81 @@ Consider the following example use case runbooks:
-  * Procedures for reviewing resource utilization and identifying opportunities for optimization using AWS Cost Explorer 
+**Related best practices:**
@@ -54 +83 @@ Consider the following example use case runbooks:
-  * Guidelines for selecting the most cost-effective Amazon EC2 instance types and purchasing models (like On-Demand, Reserved, or Spot) based on workload patterns 
+  * [ADVOPS03-BP01 Mitigate deployment risks for publisher workload changes](./advops03-bp01.html)
@@ -56 +85 @@ Consider the following example use case runbooks:
-  * Automation to right size Amazon EC2 instances, remove unused resources, and use AWS Savings Plans 
+  * [ADVOPS04-BP01 Create runbooks for the most common operational events and incidents that can impact your advertising workload](./advops04-bp01.html)
@@ -58 +86,0 @@ Consider the following example use case runbooks:
-  * Processes for periodic cost reviews and budget management 
@@ -61,0 +90 @@ Consider the following example use case runbooks:
+**Related documents:**
@@ -63 +92 @@ Consider the following example use case runbooks:
-**Data management runbook**
+  * [AWS Well-Architected Operational Readiness Review (ORR) process](https://docs.aws.amazon.com/wellarchitected/latest/operational-readiness-reviews/wa-operational-readiness-reviews.html)
@@ -65 +94 @@ Consider the following example use case runbooks:
-  * Create runbooks for: 
+  * [AWS Systems Manager runbook authoring documentation](https://docs.aws.amazon.com/systems-manager/latest/userguide/automation-documents.html)
@@ -67 +95,0 @@ Consider the following example use case runbooks:
-    * Data pipeline failures 
@@ -69 +96,0 @@ Consider the following example use case runbooks:
-    * Replication issues 
@@ -71 +97,0 @@ Consider the following example use case runbooks:
-    * Storage capacity management 
@@ -73 +99 @@ Consider the following example use case runbooks:
-    * Compliance violations 
+**Related services:**
@@ -75 +101 @@ Consider the following example use case runbooks:
-  * Include Regional considerations. 
+  * [AWS Systems Manager](https://aws.amazon.com/systems-manager/)
@@ -77 +103 @@ Consider the following example use case runbooks:
-  * Document recovery procedures. 
+  * [Amazon CloudWatch](https://aws.amazon.com/cloudwatch/)
@@ -88 +114 @@ To use the Amazon Web Services Documentation, Javascript must be enabled. Please
-ADVOPS03-BP01 Create runbooks for the most common operational events and incidents that can impact your advertising workload
+ADVOPS03-BP01 Mitigate deployment risks for publisher workload changes
@@ -90 +116 @@ ADVOPS03-BP01 Create runbooks for the most common operational events and inciden
-Manage operational processes
+ADVOPS03-BP03 Design publisher workloads to reduce defects and improve flow