System and method for restoring a database in a distributed database system
View Patent ↗A system and method is presented for database restoration in a distributed data system. The distributed data system has a first database on a first server which needs restoration. The first database is first restored with a predetermined full backup file, the full backup file being made at a first predetermined time. The first database is then restored with one or more incremental backup files, the incremental backup files being made since the first predetermined time and having the most recent backup file made at a second predetermined time. A transaction log documenting data transactions since the second predetermined time is then examined. One or more transactions listed in the transaction log after the second predetermined time are then executed since they are not included in the incremental backup files. To verify whether the restored first database has completed its restoration, a transitive vector of the first database is compared with a transitive vector of a second database on a second server of the distributed database system, wherein the second database and the restored first database are expected to be the same.
1. A method for database restoration in a distributed data system, the distributed data system having a first database on a first server which needs database restoration, the method comprising:
restoring the first database according to a predetermined full backup file, the full backup file being made at a first predetermined time;
restoring the first database with one or more incremental backup files, the incremental backup files being made since the first predetermined time and having a most recent incremental backup file of the one or more incremental backup files made at a second predetermined time;
examining a transaction log documenting data transactions since the second predetermined time;
executing, on the first database, one or more transactions listed in the transaction log after the second predetermined time; and
verifying whether the first database has completed its restoration by comparing a first transitive vector maintained in association with the first database with a second transitive vector associated with the first database, wherein the first transitive vector is maintained by a first server hosting the first database and the second transitive vector is maintained by a second server hosting the second database, and wherein the first database and the second database are synchronized if the first and second transitive vectors are the same.
2. The method of claim 1 wherein the full backup file is stored in a separate storage instrument of the first server other than one used to store the first database.
3. The method of claim 1 wherein the full backup file contains identification information to locate at least one corresponding transaction log.
4. The method of claim 1 further comprising prohibiting the first server to exchange information with any other server while restoring the first database.
5. The method of claim 1 , wherein the first transitive vector includes a state of the first database, and wherein the second transitive vector includes a state of the first database.
6. The method of claim 5 , wherein the step of verifying includes examining at least one transitive vector stored on the second server with a local synchronization vector stored on the first server.
7. The method of claim 5 , further including synchronizing the first database with the second database by generating replicas of the second database in the first database.
8. The method of claim 5 , further comprising assuring that the second server is operating before restoring the first database.
9. A system for restoring a first database, the system comprising:
a first server having a predetermined full backup file made from the first database at a first predetermined time and used for restoring the first database up to the first predetermined time;
one or more incremental backup files made since the first predetermined time and having a most recent incremental backup file of the one or more incremental backup files made at a second predetermined time; and
a transaction log documenting data transactions since the second predetermined time,
wherein the full backup file, the one or more incremental backup files, and the transaction log are stored on one or more storage devices of the first server,
wherein the first database is restored by copying the full backup file and the one or more incremental backup files, and executing, on the first database, one or more transactions listed in the transaction log after the second predetermined; and
a second server hosting a second database,
wherein verification of whether the first database has completed its restoration is made by comparing a first transitive vector maintained in association with the first database with a second transitive vector associated with the first database, wherein the first transitive vector is maintained by the first server hosting the first database and the second transitive vector is maintained by the second server hosting the second database, and wherein the first database and the second database are confirmed as synchronized if the first and second transitive vectors are the same.
10. The system of claim 9 wherein the full backup file is stored in a separate storage instrument of the first server other than one used to store the first database.
11. The system of claim 9 wherein the full backup file contains identification information to locate at least one corresponding transaction log.
12. The system of claim 9 further comprising means for prohibiting the first server to exchange information with any other server while restoring the first database.
13. The system of claim 9 , wherein the first transitive vector includes a state of the first database, and wherein the second transitive vector includes a state of the first database.
14. The system of claim 13 , further comprising means for synchronizing the first database with the second database by generating replicas of the second database in the first database.
15. The system of claim 13 , further comprising means for assuring that the second server is operating before restoring the first database.
16. A method for database restoration in a distributed data system, the distributed data system having a first database on a first server which needs database restoration, the method comprising:
assuring the first server and a second server are operating, the second server having exchanged data with the first server;
prohibiting the first server to exchange data with any other server while restoring the first database;
starting to restore the first database according to a predetermined full backup file, the full backup file being made at a first predetermined time;
restoring the first database with one or more incremental backup files, the incremental backup files being made since the first predetermined time and including a most recent incremental backup file made at a second predetermined time;
examining a transaction log documenting data transactions since the second predetermined time, the transaction log containing identification information to locate the full backup file;
executing, on the first database, one or more transactions listed in the transaction log after the second predetermined time; and
verifying whether the first database has completed its restoration by comparing a first transitive vector maintained in association with the first database with a second transitive vector associated with the first database, wherein the first transitive vector is maintained by a first server hosting the first database and the second transitive vector is maintained by a second server hosting the second database, and wherein the first database and the second database are synchronized if the first and second transitive vectors are the same.
17. The method of claim 16 , wherein the first transitive vector includes a state of the first database, and wherein the second transitive vector includes a state of the first database.