AWS Resilience Hub helps teams define recovery goals, assess application configurations and test recovery.
Application models and resiliency policies
AWS Resilience Hub helps teams assess whether an application’s recovery estimates meet business targets. In the original experience, the application model identifies resources to assess and the policy states the recovery targets.
How it works
- Import resources from CloudFormation, Terraform state, Resource Groups, EKS, or myApplications to describe the application.
- Set a recovery time objectiveThe acceptable time an application can be down after a disruption. (RTO), the acceptable downtime, and recovery point objectiveThe acceptable period of data loss after a disruption. (RPO), the acceptable period of data loss.
- Attach a policy with targets for application, infrastructure and Availability ZoneOne or more distinct data centers in an AWS Region with independent power and networking. disruptions. An Availability Zone is a distinct set of data centers in an AWS Region. Region targets are optional. Assess the model against the policy.
| Disruption | Example |
|---|---|
| Application | A required software process is lost. |
| Infrastructure | An EC2 instance is lost. |
| Availability Zone | One or more zones are unavailable. |
| Region | One or more Regions are unavailable. |
Exam cues
- If the question says a CloudFormation application needs recovery assessment, think import the stack into Resilience Hub, because its resources describe the application.
- If the question says several applications have the same recovery requirements, think a shared policy, because one policy can be attached to multiple applications.
- If the question says zero downtime is required, think a zero target is breached, because the lowest assessment result is near zero.
Remember
- A policy does not change the application.
- Region is the only disruption type whose targets may be omitted.
- The next generation became generally available on May 28, 2026. It groups services into systems, describes paths as user journeys, discovers dependencies and uses generative AI for failure analysis.
- The next generation uses modular policies for disaster recovery, availability and data recovery, and supports organization-wide governance. Existing customers can adopt it at their own pace.
Trap: Attaching a policy does not implement recovery. It sets targets for assessment.
Da documentação da AWS (19)
- Resilience Hub describes an application by importing resources from CloudFormation stacks, Terraform state files, Resource Groups, EKS clusters, or myApplications.
- A resiliency policy sets recovery time and recovery point objectives for application, infrastructure, Availability Zone and Region disruptions.
- An application disruption is the loss of a required software service or process.
- An infrastructure disruption is the loss of hardware such as an EC2 instance.
- An Availability Zone disruption makes one or more Availability Zones unavailable.
- A Region disruption makes one or more Regions unavailable.
- RTO is the maximum acceptable delay between service interruption and restoration.
- RPO is the maximum acceptable time since the last data recovery point.
- Region is the only disruption type for which RTO and RPO targets may be omitted.
- A policy does not change the application.
- Multiple applications can share a resiliency policy.
- A zero RTO or RPO target can be entered, but the lowest assessment result is near zero, so a zero target is breached.
- The next generation became generally available on May 28, 2026; existing customers may adopt it at their own pace.
- The next generation groups services into systems and uses user journeys to describe paths through them.
- The next generation uses generative AI for failure mode assessments.
- The next generation automatically discovers dependencies.
- The next generation combines disaster recovery, availability and data recovery conditions in modular policies.
- The next generation supports organization-wide governance.
- An Availability Zone is one or more distinct data centers in an AWS Region with independent power, networking and connectivity.
Assessments, drift, and pipeline checks
In the original AWS Resilience Hub experience, assessments compare recovery estimates with an application’s policy and suggest improvements. A test measures recovery during a simulated outage.
How it works
- Assess the application model against its recovery policy. Review estimated recovery time and data loss, component configuration, alarms, tests and recovery procedures. An application componentA group of related resources that work and fail as a single unit. is a group of related resources that work and fail together.
- Improve the application and reassess. Run controlled outage simulations to measure recovery.
- Enable daily assessments after the first successful manual assessment and AWS Identity and Access Management (IAM) role setup. Enable drift notifications with a permitted Amazon Simple Notification Service (SNS) topicA communication channel that sends messages to multiple endpoints., a channel for messages.
- For added or removed resources, reimport the input sources, publish the updated model and assess it.
Exam cues
- If the question says check each build before production, think Resilience Hub in the continuous integration and continuous deliveryPractices for building and testing code changes and automating the software release process. (CI/CD) pipeline, because it can validate builds before release.
- If the question says a resource was added after the model was published, think reimport, publish and assess, because the updated resources must be included.
- If the question says measure recovery under failure, think controlled outage tests, because an assessment provides estimates.
- If the question says a previously compliant component no longer meets its policy, think application policy driftA component that met its policy in the previous assessment no longer meets it in the current assessment., because the latest assessment detected a compliance change.
Remember
- Daily assessments start after a successful manual assessment and appropriate IAM role configuration.
- Policy met means all components are estimated to meet the policy; policy breached means at least one does not.
- Resource driftResources added or removed since the latest published application version. identifies added or removed resources. Policy drift identifies components that no longer meet their policy.
- Some resources are unsupported or ignored, including non-top-level resources.
Trap: A compliant estimate is not an outage test. Simulate failures to measure recovery.
Import resources and assess their configuration against recovery targets.
Run controlled failure tests to measure recovery.
Da documentação da AWS (17)
- Resilience Hub assesses application components against the policy and reports estimated workload RTO and RPO.
- An application component is a group of related resources that work and fail as a single unit.
- Assessment recommendations include component configuration, alarms, tests and recovery procedures.
- Resilience Hub can import application resources from CloudFormation stacks.
- Estimated workload RTO is based on the application configuration, not the policy.
- Tests simulate resource outages to measure resilience and recovery time.
- Daily scheduled assessment begins after a first successful manual assessment and appropriate IAM role configuration.
- Drift notifications require an SNS topic with permissions for Resilience Hub to publish.
- An SNS topic is a communication channel for messages to multiple endpoints.
- The report identifies resources added or removed since the latest published application version; reimport, publish and assess to include updated resources.
- Application policy drift identifies components that met the policy in the previous assessment but fail to meet it in the current assessment.
- Policy met means all application components are estimated to meet the policy objectives; policy breached means one or more components do not satisfy them.
- Resilience Hub can be added to a CI/CD pipeline to validate builds before production release.
- Continuous integration builds and tests code changes; continuous delivery automates the software release process.
- Some resources are unsupported or ignored, including resources that do not affect estimated RTO or RPO and non-top-level resources.
- AWS recommends controlled failure simulations with Fault Injection Service to gain confidence before production deployment.
- Resilience Hub can create and run Fault Injection Service experiments that mimic application disruptions.
Flashcards
Nada para revisar aqui agora. Escolha outros tópicos ou outro filtro.
·
Todos os 13 flashcards em lista
What targets does an AWS Resilience Hub resiliency policy contain?
RTO and RPO targets for application, infrastructure, Availability Zone and Region disruptions. Region targets are optional.
A team wants to assess resources described in a CloudFormation stack in AWS Resilience Hub. How can it model the application?
Import the stack as an application input source.
A team sets a zero RTO in an AWS Resilience Hub policy. Will a near-zero estimated RTO meet it?
No. A zero target is breached because the lowest assessment result is near zero.
Does attaching an AWS Resilience Hub policy change the application?
No. The policy supplies assessment targets and can be shared by multiple applications.
How do RTO and RPO differ in an AWS Resilience Hub policy?
RTO is acceptable downtime; RPO is the acceptable period of data loss.
A business specifies maximum acceptable downtime and data loss for a Regional outage. How should a team express these needs in AWS Resilience Hub?
Set the Region disruption RTO and RPO targets in a resiliency policy. RTO covers restoration time; RPO covers the age of the recovery point.
What does an AWS Resilience Hub assessment report?
Estimated workload RTO and RPO against policy targets, with recommendations for configuration, alarms, tests and recovery procedures.
Does an AWS Resilience Hub assessment prove an application will meet recovery targets during an outage?
No. It estimates recovery from configuration; run outage simulations to measure recovery.
How can a team detect later resource changes with AWS Resilience Hub?
Enable daily assessments and drift notifications. Notifications require a permitted SNS topic; daily assessments require an initial successful manual assessment and an appropriate IAM role.
How can a team use AWS Resilience Hub to check each build for resilience before release?
AWS Resilience Hub assessments in the CI/CD pipeline.
An AWS Resilience Hub report shows added resources after the application was published. How can the team include them in an assessment?
Reimport the input sources, publish the updated application and assess it.
A team wants to measure actual recovery after an AWS Resilience Hub assessment. What should it run?
Run controlled outage simulations, such as AWS Fault Injection Service experiments.
What does application policy drift mean in AWS Resilience Hub?
A component that met the policy in the previous assessment fails to meet it in the current assessment.