Skip to main content

EC2 (Elastic Compute Cloud)


logo

EC2 provides a scalable compute infrastructure that can be sized to match the workload’s needs.

More Information

Pricing​

On-Demand​

  • Paid by hour or seconds (depending on instance type)
  • No longer-term commitments
  • No upfront payment

Recommended for:

  • Applications with short term, spiky or unpredictable workloads that cannot be interrupted
  • Applications being developed or tested on Amazon EC2 for the first time
More Information

Spot​

  • Allows you to bid for spare EC2 computing capacity
  • Up to 90% off the On-Demand price
  • The average frequency of interruption across all Regions and instance types is < 5%
  • The billing for interrupted spot instances is different depending on whether it was you or Amazon that stopped/terminated the instance
  • Estimated savings over on-demand and frequency of interruption can be found Spot Instance advisor

Recommended for:

  • Applications that have flexible start and end times
  • Applications that are only feasible at very low compute prices
  • Users with urgent computing needs for large amounts of additional capacity
More Information

Instance Savings Plans​

  • Usage commitment (measured in $/hour) for a one or three-year period
  • All upfront, partial upfront or no upfront payment options (changes the final price)
  • Up to 72% off the On-Demand price

Recommended for:

  • Applications with steady state or predictable usage
More Information

Dedicated Hosts​

  • Dedicated physical EC2 server
  • Can be purchased on-demand (hourly) or as a a reservation (up to 70% off)
  • Allows you to use your existing server-bound licenses
  • Useful for regulatory requirements that may not support multi-tenant virtualization

Recommended for:

  • Software licenses that are bound to VMs, sockets, or physical cores
  • Corporate compliance or regulatory requirements that enforce the need to run the instances on dedicated servers instead of multi-tenant servers
More Information

Instance Types​

  • Determines the hardware of the host computer
  • Each instance type offers different compute, memory, and storage capabilities
  • Instance types are grouped in an instance family based on their capabilities.
More Information

Instance Termination​

  • When an instance terminates, the data on any instance store volumes associated with that instance is deleted
  • By default, EBS root device volumes are automatically deleted when the instance terminates, but any additional EBS volumes persists
  • You can prevent an instance from being terminated accidentally enabling Termination Protection
More Information

Instance Metadata and User Data​

Metadata​

  • Data about the instance (e.g. hostname, events, IPs)
  • http://169.254.169.254/latest/meta-data/

User Data​

  • Bootstrap configuration or script
  • By default, it runs only on first boot cycle
  • http://169.254.169.254/latest/user-data/
More Information

Placement Groups​

Placement groups influence the placement of a group of interdependent instances across underlying hardware to minimize correlated failures

  • You can't merge placement groups.
  • An instance can be launched in one placement group at a time
  • There is no charge for creating a placement group
More Information

Clustered Placement Group​

  • Grouping of instances within a single Availability Zone
  • It can span peered VPCs in the same Region
  • Not recommended for mixing instance types
  • Launching instances over time increases the chances of getting an insufficient capacity error

Recommended for:

  • Applications that benefit from low network latency or high network throughput

Partition Placement Group​

  • Amazon divides each group into logical segments called partitions
  • Each partition within a placement group has its own set of racks, which has its own network and power source
  • It can have partitions in multiple Availability Zones in the same Region

Recommended for:

  • Large distributed and replicated workloads (e.g. HDFS, HBase, Cassandra)

Spread Placement Group​

  • Group of instances that are each placed on distinct racks, with each rack having its own network and power source
  • Suitable for mixing instance types or launching instances over time
  • Can span multiple Availability Zones in the same Region

Recommended for:

  • Applications that have a small number of critical instances that should be kept separate from each other