公开(公告)号：CA2167902A1

公开(公告)日：1996-07-25

申请号：CA2167902

申请日：1996-01-23

Applicant: TANDEM COMPUTERS INC

Inventor： CARR RICHARD W ,  GARRARD BRIAN ,  MOSHER MALCOLM JR

IPC: G06F11/14 ,  G06F11/20 ,  G06F17/30

Abstract: A local computer system has a local database, application programs thatmodify the local database, and a transaction manager that stores audit records in a local audit trail reflecting those application program modifications to the local database. A remotely located computer system has a backup database. A remote data duplication facility (RDF) maintains virtual synchronization of the backup database with the local database. The RDF includes an extractor process executed by the local computer system, and a receiver process and a plurality of updater processes executed by the remote computer system. The extractor process extracts audit records from the local audit trail and transmits those records to the receiver process. The receiver distributes the received audit records to a plurality of image trail files in the remote computer system for processing by updater processes, which initiate redo operations of database modifications denoted in at least a subset of the audit records against the backup database. A catalog manager on the local computer system performs online database restructurings while application programs continue to modify the database. The transaction manager stores a Stop Updaters audit record in the local audit trail when each online database restructuring successfully completes. The extractor process transmits the Stop Updaters audit record to the remote computer system and the receiver process moves a copy each received Stop Updaters audit record into all of the image trails. Finally, each updater process stops execution when it reads a Stop Updaters audit record in its assigned image trail file.

公开(公告)号：CA2167794A1

公开(公告)日：1996-07-24

申请号：CA2167794

申请日：1996-01-22

Applicant: TANDEM COMPUTERS INC

Inventor： LYON JAMES M ,  PEARSON STEVEN R ,  PUTZOLU FRANCO ,  MOSHER MALCOLM JR

IPC: G06F11/14 ,  G06F17/30

Abstract: An improved recovery method utilizes sequence numbers to order log records and reduce the excess compensating actions due to a failure during recovery. Next undo records are written to the log after a preset number of compensating actions which include the sequence number of the record currently being scanned. After a failure, all records between the next undo record and the record having the sequence number included in the next undo record are ignored during when the log is scanned backwards.