Gravic is a world leader in providing business continuity software, enabling high and continuous availability for IT systems and payment services (including ATM and POS transactions). An inventor of bi-directional data replication, Gravic provides BASE24 Classic, BASE24-eps, Lusis Payments, and BPC SmartVista support in its patented HPE Shadowbase product line. All of the features and benefits of Shadowbase software described below are available for these four payment system environments.
Figure 1 illustrates an HPE Shadowbase active/passive replication architecture for a payment system. While users update account data on a primary node payment system (left), Shadowbase software replicates the data from a primary node to a backup node payment system (right). In the event of a primary system failure, users can fail-over to the backup system and continue processing with a current and correct version of the primary node accounts database.
Hot-site backups of production databasescan be loaded and maintained for seamless recovery in the event of a primary system failure. This Shadowbase mode is typically configured for uni-directional replication, which allows the primary node’s application to be up and running with Shadowbase data replication keeping the database on the backup node synchronized with the data changes being made on the primary node. The backup system’s application can be down ( “cold standby”) or also be up and running in a read-only mode ( “warm standby”), for example for offloading query and reporting processing from the primary system.
Figure 2 illustrates an example of an active/“almost-active” replication architecture for a payments system. While application data updates run on the active node (shown by the solid arrow), HPE Shadowbase software bi-directionally replicates the data from the primary node to the backup node payment system (shown by the dotted arrow). Test transactions can be sent to the backup node to verify processing at any point. Because the replication is bi-directional, in the event of failover, data changes will be queued in the backup node for replay when the failed node is recovered. As a result the database remains fully accessible.
HPE Shadowbase solutions also support SZT architectures where a backup node’s application is running and ready to take over on a moment’s notice. In these architectures, the application is running on both the active and the backup node, but is only accepting “update” transactions on the active node. The application on the backup node can be performing read-only query-style transactions, and can also periodically accept verification or test “update” transactions against test accounts to verify end-to-end processing. This architecture greatly improves the recovery time at failover as the target application is already running with a fully accessible database.
Figure 3 illustrates an example of an active/active replication architecture for a payments system. With the help of Shadowbase bi-directional data replication, users can be connected to and execute applications on either system (two downward solid arrows). Shadowbase bi-directional replication keeps both copies of the database synchronized, regardless of which system executes the transaction. In the event of complete failure of either active node, users connected to the failed system can be switched to the remaining system, and continue processing with no interruption. Those users originally connected to the remaining system see no failure event at all.
The Gravic patented bi-directional replication technology provides the highest level of disaster tolerance available, enabling two or more simultaneously active systems, each with its own copy of the database, within the architecture. Applications and databases can scale across multiple physical nodes to create one logical system that scales linearly and survives the failure of an entire node. HPE Shadowbase active/active technology supports various architectures that can avoid data collisions, and provide for data collision identification and resolution.
Figure 4 illustrates an example of HPE Shadowbase real-time data integration (fast data). When data is replicated from an operational system to a data warehouse using Shadowbase software, it is made available in real-time according to specifications.
Real-time data integration communicates accurate, timely data from point A to point B as the events are occurring so that better-informed business-critical decisions can be made. Application data needs integrated with sub-second latency across two or more homogeneous or heterogeneous platforms and databases. Shadowbase integration provides event-driven change data capture and replication, enabling real-time decision support as the events occur.
Seamlessly migrate to your new systems or sites, or upgrade your application or database, with minimal or zero application downtime with HPE Shadowbase Zero Downtime Migration (ZDM). A large amount of your system downtime could be due to planned outages required for upgrades, maintenance conversions, or migrations. The requirement could be for an operating system upgrade, a new application release, database maintenance, or installing new hardware. With Shadowbase ZDM, your interruptions are executed without denial of application services to end users, therefore scheduled downtime is eliminated. A location is needed either on the same node or another node (depending on the type of conversion) for the new environment. Keep your new databases synchronized with the old until all applications have been migrated to the new environment, avoiding the risk of the “big-bang” cutover approach. Shadowbase ZDM facilitates testing the new environment ahead of time too, so that the users are cutover to a known-working environment.
Feed events from BASE24, Lusis, or BPC environments into an online data warehouse or a data mart rules engine as they occur, optionally returning results to control operational decision-making processing. HPE Shadowbase solutions provide the replication and online facilities for real-time business information systems and the infrastructure to implement extreme-availability architectures.