A company is deploying a high-performance in-memory distributed database on a fleet of Amazon EC2 instances. The database nodes must synchronize data continuously with sub-millisecond network latency and maximum throughput. The database service listens on custom TCP port 7000. A Network Load Balancer (NLB) is configured to route client traffic to the database cluster on port 7000, but the NLB is currently marking all database instances as unhealthy because the target group uses the default HTTP health check configuration.
Which combination of actions will achieve the lowest node-to-node replication latency and resolve the health check issue?
- Deploy the EC2 instances in a cluster placement group within a single Availability Zone, and configure the target group health check port to 7000.Cevap
- BDeploy the EC2 instances in a spread placement group across multiple Availability Zones, and configure the target group health check port to 7000.
- CDeploy the EC2 instances in a cluster placement group within a single Availability Zone, and keep the target group health check configured for the default port 80.
- DDeploy the EC2 instances in a partition placement group across multiple Availability Zones, and configure the target group health check to use the default port 80.
Cevap
Deploying the EC2 instances in a cluster placement group within a single Availability Zone and configuring the target group health check port to 7000 is the correct solution.
Deploying the EC2 instances in a cluster placement group packs the instances close together inside a single Availability Zone, enabling low-latency, high-throughput TCP communication between nodes. Since the database application is listening on port 7000, the target group health check port must be updated to 7000 so the load balancer can verify the actual database service status, rather than checking the default port 80 where no service is listening.
Adım Adım Çözüm
Anahtar Kavram
Tightly-coupled compute architectures require cluster placement groups for sub-millisecond node-to-node latency, and target group health check ports must align with application listening ports.