System and method for providing three-way failover for a transactional database

• US 20070260696A1
• Filed: 05/02/2006
• Published: 11/08/2007
• Est. Priority Date: 05/02/2006
• Status: Active Grant

• Current Assignee: MyPoints.com, Inc. (Prodege)
• Original Assignee: MyPoints.com, Inc. (Prodege)
• Inventors: Bohannon, James, Reddy, Ananth
• Granted Patent: US 7,613,742 B2
• Time in Patent Office: Days
• US Class Current: 709/208
• CPC Class Codes: G06F 11/1662 the resynchronized component or unit being a persistent storage device re-synchronization of failed mirror storage G06F11/2082; rebuild or reconstruction of parity RAID storage G06F11/1008
• CPC Class Codes: G06F 11/2023 Failover techniques
• CPC Class Codes: G06F 11/2041 with more than one idle spare processing component
• CPC Class Codes: G06F 11/2056 by mirroring
• CPC Class Codes: G06F 11/2069 Management of state, configuration or failover
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

A method for providing three-way failover for a database server group includes identifying a master server failure or a master server shutdown, designating a first replication server as a new master server, and copying data from a second replication server to a new replication server while the first replication server functions as the new master server. The method also includes receiving new data at the server group and saving the new data at the hew master server, saving the new data in a queue at the second replication server until the set of data stored on the second replication server has been copied to the new replication server, and reading the new data saved in the queue and applying the new data to the second % replication server after the set of data has been copied to the new replication server.

Claims

1. A method for providing three-way failover for a server group, comprising:

identifying at least one of a master server failure or a master server shutdown;

designating a first replication server as a new master server;

copying a set of data from a second replication server to a new replication server while the first replication server functions as the new master server;

receiving new data at the server group and saving the new data at the new master server;

saving the new data in a queue at the second replication server until the set of data stored on the second replication server has been copied to the new replication server;

reading the new data saved in the queue and applying the new data to the second replication server after the set of data stored on the second replication server has been copied to the new replication server; and

reading the new from the new master server and applying the new data to the second replication server after the set of data stored on the second replication server has been copied to the new replication server.

2. The method of claim 1, wherein saving the new data in the queue at the second replication server comprises pulling a copy of the new data from a log at the new master server into a persistent storage queue at the second replication server.

3. The method of claim 2, wherein pulling the copy of the new data from the log at the new master server into the persistent storage queue is a first thread, and wherein reading the new data saved in a queue and applying the new data to the second replication server is a second thread.

4. The method of claim 1, further comprising saving the new data in a cache memory located at the second replication server.

5. The method of claim 1, further comprising preventing an e-mail engine from running an e-mail campaign on the second replication server until the set of data stored on the second replication server has been copied to the new replication server.

6. The method of claim 1, further comprising preventing a query from running on the second replication server until the set of data stored on the second replication server has been copied to the new replication server.

7. The method of claim 1, further comprising directing a query to the new master server until the set of data stored on the second replication server has been copied to the new replication server.

8. The method of claim 1, wherein the new data is associated with a set of existing data, the set of existing data previously stored at the first and second replication servers.

9. A method for providing three-way failover for a database server group, comprising:

determining if a master server has experienced a hardware failure or a software failure;

making a first replication server a new master server;

copying existing data stored at a second replication server to a new replication server;

receiving new data at the server group and saving the new data at the new master server, wherein the new data is associated with a portion of the existing data, the portion of the existing data previously stored at the first and second replication servers;

initiating a first thread at the second replication server until the existing data stored on the second replication server has been copied to the new replication server, wherein the first thread includes pulling a copy of the new data from a log at the new master server into a persistent storage queue at the second replication server; and

initiating a second thread at the second replication server and the new replication server after the existing data stored on the second replication server has been copied to the new replication server, wherein the second thread includes reading the new data saved in the queue and applying the new data to the second replication server and saving the new data from the new master server to the new replication server.

10. The method of claim 9, further comprising preventing an e-mail engine from running an e-mail campaign on the second replication server until the existing data stored on the second replication server has been copied to the new replication server.

11. The method of claim 9, further comprising preventing a query from running on the second replication server until the existing data stored on the second replication server has been copied to the new replication server.

12. The method of claim 9, further comprising directing a query to the new master server until the existing data stored on the second replication server has been copied to the new replication server.

13. A method for providing three-way failover for a database server group, comprising:

identifying at least one of a replication server failure or a replication server shutdown;

designating a new replication server, the new replication server operatively coupled to a first replication server, and a master server;

copying a set of data from the first replication server to the new replication server;

receiving new data at the server group and saving the new data at the master server;

saving the new data in a queue at the first replication server until the set of data stored on the first replication server has been copied to the new replication server;

reading the new data saved in the queue and applying the new data to the first replication server after the set of data stored on the first replication server has been copied to the new replication server; and

copying the new data from the new master server to the first replication server after the set of data stored on the first replication server has been copied to the new replication server.

14. The method of claim 13, further comprising preventing an e-mail engine from running an e-mail campaign on the first replication server until the set of data stored on the first replication server has been copied to the new replication server.

15. The method of claim 13, further comprising preventing a query from running on the first replication server until the set of data stored on the first replication server has been copied to the new replication server.

16. The method of claim 13, further comprising directing a query to the master server until the set of data stored on the first replication server has been copied to the new replication server.