Explanation
Correct Answer: A - Run the EC2 instances in a spread placement group.
Why this is correct:
- Spread Placement Groups are specifically designed to ensure that instances are placed on distinct underlying hardware to minimize correlated failures.
- Spread placement groups place each instance in a separate rack, with each rack having its own network and power source.
- This configuration prevents groups of nodes from sharing the same underlying hardware, which is exactly what the requirement states.
- Spread placement groups are ideal for applications that require high availability and need to minimize the risk of simultaneous failures.
Why other options are incorrect:
B. Group the EC2 instances in separate accounts.
- Separate AWS accounts do not control hardware placement. Instances in different accounts can still share underlying hardware in the same AWS region.
- This is an organizational/management solution, not a hardware isolation solution.
C. Configure the EC2 instances with dedicated tenancy.
- Dedicated tenancy ensures instances run on single-tenant hardware (dedicated to a single AWS customer), but it doesn't prevent groups of nodes from sharing the same underlying hardware with each other.
- All instances with dedicated tenancy could still be placed on the same physical server.
- This provides isolation from other AWS customers, not isolation between your own instances.
D. Configure the EC2 instances with shared tenancy.
- Shared tenancy is the default where instances can share hardware with other AWS customers.
- This is the opposite of what's needed - it actually increases the likelihood of sharing hardware.
Key AWS Concepts:
- Spread Placement Groups: Maximum of 7 instances per Availability Zone, each on distinct hardware.
- Partition Placement Groups: For large distributed and replicated workloads like HDFS, HBase, and Cassandra.
- Cluster Placement Groups: For low-latency, high-throughput workloads where instances are placed close together.
Use Case Alignment:
The requirement specifically mentions "prevent groups of nodes from sharing the same underlying hardware," which is the exact purpose of spread placement groups in AWS.
