- allow you to control the placement strategy of your EC2 instances.
- Used for high-performance computing applications and workloads requiring fast data exchange, such as big data analysis or machine learning.
- It can achieve higher throughput and lower latency because of the proximity of the instances within a single AZ.
- Has a high risk of correlated hardware failures since instances are on a single rack. Super low latency with 10GB between instances. If the rack fails, all EC2's will fail at the same time.
a) Big Data Jobs
b) Extremely low latency apps.
- Distributes instances across logical partitions to ensure that groups of instances in one partition do not share hardware with different partitions.
- Used for large distributed and replicated applications such as Hadoop, Cassandra, and data warehousing solutions.
- Each partition can contain multiple instances, and you can specify the number of partitions.
- If a partition's hardware fails, the other instances are unaffected.
- Used for applications that need high availaility.
- Maximizes availability by ensuring that resources do not share the same physical hardware, making it ideal for critical applications where minimizing downtime is crucial. - Can be used across multiple Availability Zones, making them suitable for applications that need high levels of availability.