High Availability and Load Balancing - AWS CloudHSM Classic

This is the user guide for AWS CloudHSM Classic. For the latest version, see the AWS CloudHSM User Guide.

High Availability and Load Balancing

The recommended configuration for using AWS CloudHSM Classic is to use two HSMs configured in a high-availability (HA) configuration. The failure of a single HSM appliance in a non–HA configuration can result in the permanent loss of keys and data. A minimum of two HSMs are suggested for HA purposes, with each HSM in a different Availability Zone. With this configuration, if one of your HSMs is unavailable, your keys are still available. This topic contains information about how to set up a traditional HA configuration.

            AWS CloudHSM Classic high-availability configuration

HA allows multiple HSMs to be grouped together to form one virtual device, or logical unit, as seen from the client, similar to clustering or RAID technologies. In an HA configuration, service is maintained even if one or more HSMs are unavailable. For example, if three HSMs are combined into an HA group, service is maintained even if two HSMs are offline.

When configured for HA, each HSM joins an HA group, managed through the HSM client. To HSM clients, the HA group appears as a single HSM. However, from an operational perspective, the members in the HA group share the transaction load, synchronize data with each other, and gracefully redistribute the processing capacity in the event of failure in a member HSM, to maintain uninterrupted service to the clients. HA provides load balancing across all member HSMs to increase performance and response time, while providing the assurance of HA service. All member HSMs are active (rather than one active and the rest passive). Calls are passed from each client application through the HSM client-side software (library) to one of the member HSMs on a least-busy basis.

For more information, see the Overview of Luna High Availability and Load Balancing technical note on the Gemalto SafeNet website. For more information about HA best practices, see Best Practices for High Availability and Load Balancing.

The AWS CloudHSM Classic service defines a resource known as a high-availability (HA) partition group. A high-availability partition group is a virtual partition that represents a group of partitions, typically distributed between several physical HSMs for high-availability. You use the AWS CloudHSM Classic command line interface tools to create and manage your high-availability partition groups.

            AWS CloudHSM Classic high-availability partition group configuration