Skip to content

Placement groups

This is similar to the Raid Levels ## Cluster Clusters instances into a low-latency group in a single Avalibility zone Cluster Example - Infra - Same Rack(same hardwere) - Same AZ(avalibility zone) - Pros - Greate network(10Gbps bandwith between instances) - Cons - If the rack fails, all instances fail.

  • Use case:
    • Big Data Job that need to complete fast
    • App that needs extremely low latency and high network throughput

Spread

Spreads instances across underlying hardwere (Max 7 instace per group per AZ) #CriticalApps Spread Example

  • Pros
    • Can span accross AZ
    • Reduce simultaneous failure
  • Cons
    • Only 7 instances
  • Use Case
    • Aplication that maximzie high avalibility
    • Each instance has to be isolated form faileur from each other ### Can Operate on racks and hosts Aws host outputs

Partition

Example Spreads accrros many diffrent partiction (relay on diffrent sets of racks) - Scales to 100 EC2 instaces per gorup (Kafka,Cassandra) - a partition failure can affect many EC2 but won’t affect otehr partitions - EC2 instance get access to the partition metadata - Use cases
- Kafka, Cassandra

Ec2instance