It is the strategy that we place the EC2 instance.
There are 3 strategies
Cluster
Spread
Partition
It will places all EC2 instances in 1 AZ (Single AZ).
Low network latency
Low High availability because If the AZ fails , all instances will fail.
Big data needs to complete fast
Application need low latency
It will places all instances on serparate AZ.
can span across AZ
reduce risk failure
limited 7 instance per AZ
application that needs to maximize high availablity
It is similarto Spread but It can span across multiple AZ in the same region. The instance in a partion do not shre racks with other partition.
can span multiple AZ in same region
one partition fail all of instance in that partition will fail
Kafka, Cassandra
Last updated 1 year ago