Skip to main content

Products & services  >  Software  >  Database and Data Management
IBM RepliData  

More information
RepliData opens up a new avenue of opportunity for organizations considering multi-site backup and/or operation of production database applications. The key benefits include: High performance

Production experience at a customer site shows that RepliData can process change activity of 94 transactions per second, with four changed rows per transaction, with latency of less than 5 seconds, using the equivalent of 7 per cent of a 9672-RX5. This is equivalent to over 1.3 million database changes per hour.

Low latency

RepliData provides a database replication engine that can apply changes in replica databases in near real-time. This is partly accomplished through RepliData's architecture which is a 'push' rather than a 'pull' operation. The limiting factors on RepliData replication latency are the amount of CPU resource available on the source and target systems, and the amount of network bandwidth available for transmission. The package creation, package transmission, and package apply processes are all done as multiple, parallel operations so that they can take advantage of all resources available.

Data integrity

RepliData creates integral packages of changes, where a package contains all the changes successfully committed at the source in a single unit of work. The package is transmitted as a single entity to the target systems. The RepliData apply process applies all the changes in the package as a single unit of work, such that all changes are successfully applied or none of them are. In either case, the target database retains internal consistency.

Peer-to-peer replication

RepliData supports 'peer-to-peer' replication. In this mode of operation, database changes can originate at all systems concurrently. However, RepliData contains a collision-detection function that recognises changes made to the same row on different systems. When a 'collision' is encountered, RepliData ensures that only the most current changes are applied.

Conditional scenarios

RepliData has an intelligent replication engine which can distinguish between different scenarios and can selectively replicate data to different sites as a result.

Industrial strength

RepliData is being used within parts of IBM, as well as by a number of customers, for complex business-critical applications involving high-volume database replication across multiple sites. These include major financial, post and telecommunications institutions in the US and Europe.

IBM uses RepliData for some of its own key business applications including one which entails database replication between sites in the US and Europe.