Amazon AWS Certified Solutions Architect - Professional SAP-C02 Bundle

This alternative, Migrate eligible tasks to ECS with Fargate launch type, is appropriate when A containerized API already has an ECS task definition and team wants to stop managing node instances. Fargate removes host provisioning while retaining ECS task and service concepts.

For the separate situation A legacy application has a tightly coupled OS service and requires privileged host-level changes, Keep it on managed EC2 instances until those OS dependencies can be removed could be useful. EC2 retains operating-system control unavailable in fully managed serverless services.

Although Move the handler to Lambda after validating runtime and request constraints can be valid when A migrated HTTP handler is stateless, runs for seconds and has unpredictable demand, it targets a different technical condition. Lambda handles short on-demand work without persistent server capacity.

Bind the workload service account to an appropriately scoped EKS Pod Identity role solves the related scenario A Kubernetes workload needs access to one S3 bucket but should not inherit node IAM permissions; workload-specific IAM identities avoid excessive shared node access.

Select AWS Fargate capacity for eligible ECS services would address a different requirement: A small team wants managed compute for ECS containers with no EC2 instance fleet. Fargate runs compatible containers without customer-maintained worker hosts.

This alternative, Pin its task or Pod image reference to an immutable ECR digest, is appropriate when A container service must deploy the exact version tested in staging and avoid tag drift. Immutable image references preserve application provenance across deployment stages.

Although Use Amazon EKS for managed Kubernetes control-plane support can be valid when A microservice estate already depends on upstream Kubernetes APIs and operators, it targets a different technical condition. EKS retains Kubernetes compatibility while AWS operates the control plane.

This alternative, Select AWS Fargate capacity for eligible ECS services, is appropriate when A small team wants managed compute for ECS containers with no EC2 instance fleet. Fargate runs compatible containers without customer-maintained worker hosts.

For the separate situation A container service must deploy the exact version tested in staging and avoid tag drift, Pin its task or Pod image reference to an immutable ECR digest could be useful. Immutable image references preserve application provenance across deployment stages.

This alternative, Run it on Amazon ECS with suitable EC2 or Fargate capacity, is appropriate when A service uses Docker images but has no Kubernetes-specific requirements and prioritizes AWS-native simplicity. ECS provides managed container orchestration without Kubernetes control-plane complexity.

For the separate situation A Kubernetes workload needs access to one S3 bucket but should not inherit node IAM permissions, Bind the workload service account to an appropriately scoped EKS Pod Identity role could be useful. Workload-specific IAM identities avoid excessive shared node access.

Use Amazon EKS for managed Kubernetes control-plane support solves the related scenario A microservice estate already depends on upstream Kubernetes APIs and operators; eKS retains Kubernetes compatibility while AWS operates the control plane.

This alternative, Bind the workload service account to an appropriately scoped EKS Pod Identity role, is appropriate when A Kubernetes workload needs access to one S3 bucket but should not inherit node IAM permissions. Workload-specific IAM identities avoid excessive shared node access.

For the separate situation A microservice estate already depends on upstream Kubernetes APIs and operators, Use Amazon EKS for managed Kubernetes control-plane support could be useful. EKS retains Kubernetes compatibility while AWS operates the control plane.

Although Select AWS Fargate capacity for eligible ECS services can be valid when A small team wants managed compute for ECS containers with no EC2 instance fleet, it targets a different technical condition. Fargate runs compatible containers without customer-maintained worker hosts.

Run it on Amazon ECS with suitable EC2 or Fargate capacity solves the related scenario A service uses Docker images but has no Kubernetes-specific requirements and prioritizes AWS-native simplicity; eCS provides managed container orchestration without Kubernetes control-plane complexity.

Although Select AWS Fargate capacity for eligible ECS services can be valid when A small team wants managed compute for ECS containers with no EC2 instance fleet, it targets a different technical condition. Fargate runs compatible containers without customer-maintained worker hosts.

Pin its task or Pod image reference to an immutable ECR digest would address a different requirement: A container service must deploy the exact version tested in staging and avoid tag drift. Immutable image references preserve application provenance across deployment stages.

Select EFS with supported access points and POSIX authorization would address a different requirement: A migrated Linux application requires a shared POSIX file system mounted concurrently by hundreds of workers. EFS presents managed shared NFS rather than isolated host block volumes.

This alternative, Move the photos to S3 object storage and expose them through application or CDN URLs, is appropriate when An object library containing product photos is accessed through HTTP URLs and has no filesystem mount requirement. S3 separates object persistence from host-attached file systems.

For the separate situation An HPC migration requires parallel shared high-throughput files for simulation output, Use FSx for Lustre with a deployment mode aligned to throughput and durability could be useful. Lustre supports parallel compute access patterns that generic file shares may not satisfy.

Move the photos to S3 object storage and expose them through application or CDN URLs would address a different requirement: An object library containing product photos is accessed through HTTP URLs and has no filesystem mount requirement. S3 separates object persistence from host-attached file systems.

This alternative, Use FSx for Lustre with a deployment mode aligned to throughput and durability, is appropriate when An HPC migration requires parallel shared high-throughput files for simulation output. Lustre supports parallel compute access patterns that generic file shares may not satisfy.

For the separate situation A transactional database on EC2 needs low-latency persistent block storage, Attach appropriately sized EBS SSD volumes and test IOPS and throughput could be useful. EBS provides persistent EC2 block devices suited to database files.

For the separate situation An object library containing product photos is accessed through HTTP URLs and has no filesystem mount requirement, Move the photos to S3 object storage and expose them through application or CDN URLs could be useful. S3 separates object persistence from host-attached file systems.

Migrate shares to Amazon FSx for Windows File Server with compatible directory configuration solves the related scenario A migrated Windows file server must preserve SMB paths and Active Directory access semantics; fSx for Windows supports SMB and Windows ACL expectations.

Although Select EFS with supported access points and POSIX authorization can be valid when A migrated Linux application requires a shared POSIX file system mounted concurrently by hundreds of workers, it targets a different technical condition. EFS presents managed shared NFS rather than isolated host block volumes.

Migrate shares to Amazon FSx for Windows File Server with compatible directory configuration solves the related scenario A migrated Windows file server must preserve SMB paths and Active Directory access semantics; fSx for Windows supports SMB and Windows ACL expectations.

Although Select EFS with supported access points and POSIX authorization can be valid when A migrated Linux application requires a shared POSIX file system mounted concurrently by hundreds of workers, it targets a different technical condition. EFS presents managed shared NFS rather than isolated host block volumes.

Use FSx for Lustre with a deployment mode aligned to throughput and durability would address a different requirement: An HPC migration requires parallel shared high-throughput files for simulation output. Lustre supports parallel compute access patterns that generic file shares may not satisfy.

Attach appropriately sized EBS SSD volumes and test IOPS and throughput solves the related scenario A transactional database on EC2 needs low-latency persistent block storage; eBS provides persistent EC2 block devices suited to database files.

Although Move the photos to S3 object storage and expose them through application or CDN URLs can be valid when An object library containing product photos is accessed through HTTP URLs and has no filesystem mount requirement, it targets a different technical condition. S3 separates object persistence from host-attached file systems.

Migrate shares to Amazon FSx for Windows File Server with compatible directory configuration would address a different requirement: A migrated Windows file server must preserve SMB paths and Active Directory access semantics. FSx for Windows supports SMB and Windows ACL expectations.