A genomics research institute is deploying a DNA sequencing pipeline on Amazon EC2 instances. The workload requires a tightly-coupled cluster of compute nodes to perform distributed sequence alignment using Message Passing Interface (MPI). In addition, a web dashboard queries a status daemon running on port of these instances to monitor alignment progress. The dashboard traffic is distributed to the instances via an Application Load Balancer (ALB) listening on port .
Which TWO configurations should a solutions architect implement to achieve optimal network performance for the MPI workload and ensure the ALB correctly monitors the status daemon?
- Deploy the EC2 instances in a cluster placement group within a single Availability Zone.Cevap
- BDeploy the EC2 instances in a spread placement group across multiple Availability Zones.
- Configure the ALB target group's health checks to use port instead of the default traffic port.Cevap
- DConfigure the ALB target group to perform health checks on the default traffic port of .
- EDeploy the EC2 instances in a partition placement group to isolate processing workloads on distinct hardware partitions.
Cevap
Deploy the EC2 instances in a cluster placement group within a single Availability Zone, and configure the Application Load Balancer target group's health checks to use the custom daemon port instead of the default traffic port.
The correct configurations are to deploy the EC2 instances in a cluster placement group and to configure the ALB target group's health checks to use the specific port where the status daemon runs. A cluster placement group places instances physically close to each other in a single Availability Zone, providing the low-latency network performance required by tightly-coupled MPI workloads. Additionally, specifying the custom daemon port for the ALB health checks ensures the load balancer accurately monitors the status daemon's health rather than the main web server port.
Adım Adım Çözüm
Anahtar Kavram
Tightly-coupled high-performance computing (HPC) environments require cluster placement groups for low-latency networking, and load balancers must have health checks aligned with the specific service ports of monitored daemons.