Exam AWS Certified SysOps Administrator - Associate topic 1 question 122 discussion

CD D. Amazon ALWAYS touts the "cluster placment group" as being the solution HPC workloads (low latency between nodes). However this actually precludes resilience through ELB and multi-AZ spread of instances or multiple partitions (insufficient throughput). C. The only option I see to get both resilience and HPC is to rely on auto-scaling group to re-scale for any failing nodes. Provides some protection unless the whole rack fails. No, it's not multi-AZ or multi rack, but its better than nothing given Amazons HPC recommendations.

To meet the requirement of minimum latency between nodes for a high performance computing (HPC) application on a resilient tier of Amazon EC2 instances, the SysOps administrator should take the following actions: Launch the EC2 instances into a cluster placement group to ensure that the instances are placed in a low-latency, single-tenant infrastructure. Place the EC2 instances in an Auto Scaling group within a single subnet. This will ensure that the instances are in the same Availability Zone and will not have to traverse a network boundary to communicate with each other.

D: https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/placement-groups.html

C and D. An Auto Scaling group will provide the resilience and the Cluster placement group will provide the low latency. https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/placement-groups.html

B and E

C. Place the EC2 instances in an Auto Scaling group within a single subnet. D. Launch the EC2 instances into a cluster placement group.

I was going for B and D too, but here's the thing with B: Cluster placement groups are deployed within AN availability zone, meaning the multi-az LB does not apply: https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/placement-groups.html

E - you need resilient tier, which partition placement groups provide, not C because it doesn't (although it provides minimum latency between nodes). B - you need LB with low latency to distribute traffic between your instances, placed in partition placement groups.

Cluster – packs instances close together inside an Availability Zone. This strategy enables workloads to achieve the low-latency network performance necessary for tightly-coupled node-to-node communication that is typical of high-performance computing (HPC) applications. Ref link: https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/placement-groups.html

ADADAD

I vote for A & D. Question is about low latency and HPC, therefore D is the right placement option (so, no E). Furthermore, cluster placement will use a single availability zone, hence multi-AZ NLB is useless (so, no B). Finally, regarding ASG in a single subnet isn't resilient. Below I paste directly from AWS docs: "If you try to add more instances to the placement group later, or if you try to launch more than one instance type in the placement group, you increase your chances of getting an insufficient capacity error." So in case of a failed health check, ASG might try to spin up a new instance but receive an InsufficientInstanceCapacity (which is an AZ-specific error). Therefore, is not recommended when resiliency is needed. The only thing left is A. PS. Not a very good question, I believe a bit more context is required.

B for resilience https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/disaster-recovery-resiliency.html D for minimum latency https://docs.aws.amazon.com/AWSEC2/latest/UserGuide/placement-groups.html

Place the EC2 instances in an Auto Scaling group within a single subnet. Launch the EC2 instances into a partition placement group. Not B While Multi-AZ deployment offers high availability, some workloads are more sensitive to internode latency and could not be deployed across multiple zones. With partition placement groups, you can now deploy these workloads within a single zone and reduce the likelihood of correlated failures, improving your application performance and availability. Not D Customers wanted a way to reduce correlated failures for large distributed and replicated workloads that required hundreds of EC2 instances. To isolate the impact of hardware faults, EC2 subdivides each partition placement group, into logical segments called partitions. EC2 ensures that no two partitions within a placement group share the same racks. https://aws.amazon.com/blogs/compute/using-partition-placement-groups-for-large-distributed-and-replicated-workloads-in-amazon-ec2/

BE Resilient is not 1 subnet. Hence a network load balancer.

CDCDCDCDCD

C and D

C and D