# US 6,732,123 B1 · Database recovery to any point in time in an online environment utilizing disaster recovery technology

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 6,732,123 B1
- Document ID: US6732123B1
- Status: Granted
- Kind: B1
- Application number: 09/028,193
- Filed: 1998-02-23
- Granted: 2004-05-04
- Earliest claimed: 1998-02-23
- Assignees: International Business Machines Corporation
- Inventors: David Wayne Moore; Karen Alicia Ranson; Lawrence Edward Sullivan; Donald Prentiss Terry; Gregory Walter Vance; Vern Lee Watts
- Web page: https://patentgrep.com/patent/US6732123B1

## CPC

- G06F 11/1471: involving logging of persistent data for recovery
- G06F 2201/80: Database-specific techniques
- Y10S 707/99952: Coherency, e.g. same view to multiple users
- Y10S 707/99953: Recoverability
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- Y10S: TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS

## Abstract

A method and apparatus for merging log data for recovering one or more database data sets to any point-in-time in a single pass of log data sets. Using disaster recovery technology, the invention may be implemented to provide a method to reduce the time that a broken database is unavailable by processing recovery input data in parallel and recovering multiple database data sets, including areas, simultaneously. The invention reduces operational complexity and resource constraints experienced by IMS customers trying to recover database data sets especially in a data sharing environment. The invention simplifies the recovery process by eliminating both the need for change accumulation and the need to run separate recovery jobs for each database data set requiring recovery, thereby reducing the impact of recovery on system resources.

## Claims

### Claim 1

A method for recovering multiple databases in an information management system, the method comprising:
maintaining log records in parallel from the multiple logs;
in a single pass:
reading log records in parallel from the multiple logs; and
merging the log records in a single log stream in creation time sequence, creation time being the time a log record is entered into a log; and
recovering the multiple databases substantially simultaneously using the single log stream wherein the amount of time a broken database is unavailable is substantially reduced.

### Claim 2 (depends on 1)

The method recited in claim 1, wherein:
recovering includes using only log records associated with completed units of recovery.

### Claim 3 (depends on 2)

The method recited in claim 2, further comprising:
accepting a recovery time, the recovery time indicating a point in time at which recovery of a database will end; and
recovering includes using log records until the recovery time is reached.

### Claim 4 (depends on 3)

The method recited in claim 3, the method further comprising:
reading log data sets past the recovery time;
determining whether a database data set not involved in recovery exists in the log;
determining whether a unit of recovery not completed during the recovery time updated a database data set not involved in the recovery; and
notifying that database data sets not being recovered are potentially inconsistent with database data sets being recovered.

### Claim 5 (depends on 4)

The method recited in claim 4, wherein the recovery time does not occur at an end of an allocation boundary.

### Claim 6 (depends on 5)

The method recited in claim 5, the method including backing-out all updates associated with any unit of recovery that has not completed-prior to the recovery time.

### Claim 7

A method for recovering databases in an information management system using disaster recovery technology, the method comprising:
maintaining multiple databases, and multiple logs, each log having log records;
in a single pass:
reading log records in parallel from the multiple logs;
merging the log records into a single log stream in creation time sequence, creation time being the time a log record is entered into a respective log; and
recovering the multiple databases substantially simultaneously using the single log stream wherein the amount of time a broken database is unavailable is substantially reduced.

### Claim 8 (depends on 7)

The method recited in claim 7, further comprising:
accepting a recovery time, the recovery time indicating a point in time at which recovery of a database will end; and
recovering includes using log records until the recovery time is reached.

### Claim 9 (depends on 1)

The method recited in claim 1, the method further comprising:
reading log data sets past the recovery time;
determining whether a database data set not involved in recovery exists in the log;
determining whether a unit of recovery not completed during the recovery time updated a database data set not being recovered; and
notifying that database data sets not being recovered are potentially inconsistent with database data sets being recovered.

### Claim 10 (depends on 9)

The method recited in claim 9, wherein the recovery time does not occur at an end of an allocation boundary.

### Claim 11 (depends on 10)

The method recited in claim 10, the method including backing-out all updates associated with any unit of recovery that has not completed prior to the recovery time.

### Claim 12

A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform a method for merging log records for recovering a database, the database including one or more databases, said method comprising:
maintaining log records for the multiple databases at multiple logs;
in a single pass:
reading log records in parallel from the multiple logs; and
merging the log records into a single log stream in creation time sequence, creation time being the time a log record is entered into a log; and
recovering the multiple databases substantially simultaneously using the single log stream wherein the amount of time a broken database is unavailable is substantially reduced.

### Claim 13 (depends on 12)

The signal bearing medium recited in claim 12, the method further comprising:
recovering includes using only log records associated with completed units of recovery.

### Claim 14 (depends on 13)

The signal bearing medium recited in claim 13, the method further comprising:
accepting a recovery time, the recovery time indicating a point in time at which recovery of a database will end; and
recovering includes using log records until the recovery time is reached.

### Claim 15 (depends on 14)

The signal bearing medium recited in claim 14, the method further comprising:
reading log data sets past the recovery time;
determining whether a database data set not involved in recovery exists in the log;
determining whether a unit of recovery not completed during the recovery time updated a database data set not involved in recovery; and
notifying that further database data sets not being recovered are potentially inconsistent with database data sets being recovered.

### Claim 16 (depends on 15)

The signal bearing medium recited in claim 15, the method further comprising backing-out all updates associated with any unit of recovery that has not completed prior to the recovery time.

### Claim 17

An information management system to manage merging of log records to recover data, the system comprising:
multiple databases;
storage including multiple logs, each log having log records;
a processor communicatively coupled to the multiple databases and storage, the processor executing commands to recover a database by:
maintaining log records for the multiple databases at the multiple logs;
in a single pass:
reading log records in parallel from the multiple logs; and
merging the log records into single log stream in creation time sequence, creation time being the time a log record is entered into a log; and
recovering the multiple databases substantially simultaneously using the single log stream wherein the amount of time a broken database is unavailable is substantially reduced.

### Claim 18 (depends on 17)

The system recited in claim 17, wherein:
recovering includes using only log records associated with completed units of recovery.

### Claim 19 (depends on 18)

The system recited in claim 18, the commands executed by the processor further comprising:
accepting a recovery time, the recovery time indicating a point in time at which recovery of a database will end; and
recovering includes using log records until the recovery time is reached.

### Claim 20 (depends on 19)

The system recited in claim 19, the commands executed by the processor further comprising:
reading log data sets past the recovery time;
determining whether a database data set not involved in recovery exists in the log;
determining whether a unit of recovery not completed during the recovery time updated a database data set not being recovered; and
notifying that further database data sets not being recovered are potentially inconsistent with database data sets being recovered.

### Claim 21 (depends on 20)

The system recited in claim 20, the commands executed by the process or further including backing-out all updates associated with any unit of recovery that has not completed prior to the recovery time.

### Claim 22

An information management system to manage merging of log records to recover data, the system comprising:
storage means for storing data;
multiple log means for storing updates to data stored in the data means;
processor means communicatively coupled to the storage means and the log means, the processor means executing commands to recover a database by:
maintaining log records for the multiple databases at the log means;
in a single pass:
reading log records in parallel from the log means; and
merging the log records into a single log stream in creation time sequence, creation time being the time a log record is entered into a log; and
recovering the multiple databases substantially simultaneously using the single log stream wherein the amount of time a broken database is unavailable is substantially reduced.

### Claim 23 (depends on 22)

The system recited in claim 22, wherein:
recovering includes using only log records associated with completed units of recovery.

### Claim 24 (depends on 23)

The system recited in claim 23, the processor executing further commands to recover a database comprising:
accepting a recovery time, the recovery time indicating a point in time at which recovery of a database will end, wherein the recovery time does not occur at an allocation boundary; and
recovering includes using log records until the recovery time is reached.

### Claim 25 (depends on 24)

The system recited in claim 24, the processor executing further commands to recover a database comprising:
reading log data sets past the recovery time; and
determining whether further log data sets remain in the log.

### Claim 26 (depends on 25)

The system recited in claim 25, the processor executing further commands to recover a database comprising:
determining whether a database data set not involved in recovery exists in the log;
determining whether a unit of recovery not completed during the recovery time updated the database data set not being recovered; and
notifying that database data sets not being recovered are potentially inconsistent with database data sets being recovered.

### Claim 27 (depends on 26)

The system recited in claim 26, the processor executing further commands to recover a database comprising saving the further data sets.

### Claim 28 (depends on 27)

The system recited in claim 27, the processor executing commands to recover a database including backing-out all data sets associated with any transaction that has not completed prior to the recovery time.
