July 18, 2013 | abalogh
Ensuring High Availability with InfiniDB
InfiniDB is used for high-value, Big Data analytics deployments where processing speed is essential. The ability to analyze tens or hundreds of Terabytes and return results in seconds is often utilized for mission-critical Business Intelligence and advanced analytics applications.
As we work with customers that design and build these kinds of applications, we are often asked “what High Availability options are available for InfiniDB?”
Fortunately, InfiniDB was architected specifically for large-scale deployments and has many built-in capabilities for redundancy and high availability. Its two main processing components, the User Module and Performance Module, natively handle redundancy and failover. As with most systems of this type, however, what InfiniDB does internally is only part of an overall HA strategy. An application architecture using InfiniDB will also need to address many other types of redundancy (network/IP redundancy, storage/disk redundancy, power redundancy, etc.) to implement an effective HA strategy.
As we help customers evaluate options for high-availability deployment, there are some of the aspects we consider:
- What are the availability requirements for your application?
- What are the performance requirements for your application?
- Will the deployment be on-premise or cloud-based?
- What platform infrastructure already exists (networking, storage, etc.)?
- How does your application expect/need to interface to the storage engine?
For full details and more information on ensuring High Availability in your InfiniDB cluster, please see our InfiniDB High Availability Guide.