AWS EC2 – Placement Groups – Certification

Placement Groups Overview

  • A Placement Group is a logical grouping of instances within a single Availability Zone and are recommended for applications that benefits from low network latency, high network throughput, or both.
  • Placement group don’t span across Availability Zones
  • Placement group is only available within a single Availability Zone either in the same VPC or peered VPCs
  • Placement group is more of an hint to AWS that the instances need to be launched physically close to each together
  • Using placement groups enables applications to participate in a low-latency, 10 Gbps network.

AWS EC2 Placement Group

Placement Group Key Points

  • Provides Low network latency & high network throughput and is recommened for HPC (High Performance Computing)
  • Should be in the same Availability Zone and can’t span multiple AZs
  • Should have unique name within AWS account
  • To provide the lowest latency, and the highest packet-per-second network performance, choose instance type that supports enhanced networking
  • Only Specific Instance types (General Purpose, GPU, Compute, Memory, Storage Optimized – m4.10xlarge, c4.8xlarge, c3.8xlarge, g2.8xlarge, r3.8xlarge, i2.8xlarge, d2.8xlarge)  which support 10 Gigabyte network performance can take full advantage of Placement Group
  • Placement Groups cannot be merged. Instances in one placement group should be terminated, and then relaunched into an other placement group
  • Existing EC2 instance can’t be moved into a Placement Group as it might be physically located away from the existing instances
  • For moving an instance into the placement group,
    • create an AMI from the existing instance,
    • and then launch a new instance from the AMI into a placement group.
  • AWS recommends using the same homogenous instance type for all instances in a placement group
  • Placement group can span peered VPCs
  • When launching an instance in the placement group results in a capacity error, stop and start the instances in the placement group and try to launch the instance again

Placement Groups Best Practices

  • Use homogenous instance types
  • Launch all the placement group instances at the same time
  • Not a best fit for horizontally scalable web services
  • Ensure there is enough capacity
  • Jumbo frames help to achieve the maximum network throughput possible, and they are recommended for instances collocated inside a placement group

AWS Certification Exam Practice Questions

  • Questions are collected from Internet and the answers are marked as per my knowledge and understanding (which might differ with yours).
  • AWS services are updated everyday and both the answers and questions might be outdated soon, so research accordingly.
  • AWS exam questions are not updated to keep up the pace with AWS updates, so even if the underlying feature has changed the question might not be updated
  • Open to further feedback, discussion and correction.
  1. What is a placement group?
    • A collection of Auto Scaling groups in the same Region
    • Feature that enables EC2 instances to interact with each other via high bandwidth, low latency connections
    • A collection of Elastic Load Balancers in the same Region or Availability Zone
    • A collection of authorized Cloud Front edge locations for a distribution
  2. In order to optimize performance for a compute cluster that requires low inter-node latency, which feature in the following list should you use?
    • AWS Direct Connect
    • Placement Groups
    • VPC private subnets
    • EC2 Dedicated Instances
    • Multiple Availability Zones
  3. What is required to achieve gigabit network throughput on EC2? You already selected cluster-compute, 10GB instances with enhanced networking, and your workload is already network-bound, but you are not seeing 10 gigabit speeds.
    1. Enable biplex networking on your servers, so packets are non-blocking in both directions and there’s no switching overhead.
    2. Ensure the instances are in different VPCs so you don’t saturate the Internet Gateway on any one VPC.
    3. Select PIOPS for your drives and mount several, so you can provision sufficient disk throughput
    4. Use a placement group for your instances so the instances are physically near each other in the same Availability Zone. (You are not guaranteed 10 gigabit performance, except within a placement group. Using placement groups enables applications to participate in a low-latency, 10 Gbps network)
  4. You need the absolute highest possible network performance for a cluster computing application. You already selected homogeneous instance types supporting 10 gigabit enhanced networking, made sure that your workload was network bound, and put the instances in a placement group. What is the last optimization you can make?
    1. Use 9001 MTU instead of 1500 for Jumbo Frames, to raise packet body to packet overhead ratios. (For instances that are collocated inside a placement group, jumbo frames help to achieve the maximum network throughput possible, and they are recommended in this case)
    2. Segregate the instances into different peered VPCs while keeping them all in a placement group, so each one has its own Internet Gateway.
    3. Bake an AMI for the instances and relaunch, so the instances are fresh in the placement group and do not have noisy neighbors
    4. Turn off SYN/ACK on your TCP stack or begin using UDP for higher throughput.

References

EC2_User_Guide – Placement_Groups

6 thoughts on “AWS EC2 – Placement Groups – Certification

  1. I have a question on Placement Group. We know all instances must be placed in the same AZ for the PG to work. But AWS docs says that a PG can span peered VPCs. My question is that if peered VPCs are involved then it means a different AZs.

    I mean ….if VPCs peer with each other, then they both have different IP segments. They cannot have overlapping CIDRs. Therefore their subnets and AZs are different from each other. So, how can PG span across two peered VPCs?

    1. Placement group allows instances to be launched within it and cannot span across AZ.
      It works for VPC peered as well, however the instances launched within the peered VPC should also be in the same AZ to be added to the placement group.
      It cannot span across AZ even in different VPC.

  2. Hi Jayendra,

    >>Placement group is only available for One VPC and One Availability Zone
    Placement groups can span peered VPCs. So I guess the above statement should read “Placement group is only available with in a single Availability Zone and across peered VPCs”, right?

    Cheers,
    Satish

Leave a Reply

Your email address will not be published. Required fields are marked *