Abstract: A method and system for purging database update image files after completion of associated transactions for a database replication system with multiple audit logs. Audit update records and audit backout records are generated by the primary system, and are transmitted to the backup system in multiple streams in parallel. The backup system stores the received audit records as audit image trails, and applies the audit updates and audit backouts to the backup database without regard to whether the associated transactions committed or aborted. Because audit updates and audit backouts are applied without regard to whether the associated transactions committed or aborted, image files containing audit records associated with transactions that have not yet committed or aborted must not be purged. The present invention provides for a method for determining whether an image trail file contains audit records that can be purged such that image trail files can be safely deleted.

Abstract: A remote data facility (RDF) capable of performing a lockstep data replication procedure (“LockStep Procedure”). When the LockStep Procedure is invoked, and when an application program has committed a transaction, the application program is prevented from executing other procedures until the application is notified that audit records associated with that transaction have been safely stored to the backup system. Since the application program is prevented from executing other procedures, no decision based on the commit will be made until after the application is notified that all the audit records associated with the transaction are safely stored in the backup system.

Abstract: A method and system for purging database update image files after completion of associated transactions for a database replication system with multiple audit logs. Audit update records and audit backout records are generated by the primary system, and are transmitted to the backup system in multiple streams in parallel. The backup system stores the received audit records as audit image trails, and applies the audit updates and audit backouts to the backup database without regard to whether the associated transactions committed or aborted. Because audit updates and audit backouts are applied without regard to whether the associated transactions committed or aborted, image files containing audit records associated with transactions that have not yet committed or aborted must not be purged. The present invention provides for a method for determining whether an image trail file contains audit records that can be purged such that image trail files can be safely deleted.

Abstract: A method and system for high-speed database replication. Audit update records and audit backout records are generated by the primary system, and are transmitted to the backup system in multiple streams in parallel. The backup system stores the received audit records as audit image trails, and applies the audit updates and audit backouts to the backup database without regard to whether the transactions committed or aborted and without regard to whether the backup system received a complete set of the audit records pertaining to the transactions. Upon the occurrence of a predetermined event, the backup system applies all the audit updates and backouts it received, and subsequently “undoes” questionable audit updates and audit backouts.

Abstract: A method and system for backing up primary nodes onto backup nodes where the primary nodes can each originate a distributed transaction and can participate in a distributed transaction. The backup nodes, after a primary node failure, undo all transactions whose state is unknown on the backup node and all committed transactions which, if kept, would lead to an inconsistent set of backup nodes. First, committed transactions are undone if commits for them were not received on all back up nodes to the primary nodes that participated in the transaction. Second, all committed transactions that potentially depend on the committed transactions that were undone are also undone. Only those transactions that are provably independent of the committed transactions that were undone are kept. The result is a set of back up nodes which can support operations until the failing primary node is restored.