Automatically Performing Failover Operations With A Load Balancer

• US 20130007505A1
• Filed: 06/30/2011
• Published: 01/03/2013
• Est. Priority Date: 06/30/2011
• Status: Active Grant

• Current Assignee: Rackspace US Incorporated (Apollo Global Management, Inc.)
• Original Assignee: Rackspace US Incorporated (Apollo Global Management, Inc.)
• Inventors: Spear, Thomas D.
• Granted Patent: US 8,661,287 B2
• Time in Patent Office: Days
• US Class Current: 714/4.11
• CPC Class Codes: G06F 11/2025 using centralised failover control functionality
• CPC Class Codes: G06F 11/2038 with a single idle spare processing component

Abstract

A load balancer includes a failover logic unit to identify servers to execute services, generate and store in the load balancer a failover rule and a service rule, and to determine a failure in a first server that executes a first service responsive to a lack of response by the first server to a keepalive message sent by the load balancer to the first server. The load balancer can then perform an operation to cause an automatic failover of the first service to another server based on the failover and service rules.

Claims

1. A method comprising:

receiving, in a load balancer, a status message from each of a plurality of services operating on a plurality of servers coupled behind the load balancer;

detecting a failure of a first service executing on a first server of the plurality of servers based on a status message; and

automatically enabling a second server corresponding to a failover server to the first server to execute the first service, via the load balancer.

2. The method of claim 1, wherein the plurality of servers comprises a cluster.

3. The method of claim 1, further comprising reporting the failure via an outside monitoring entity.

4. The method of claim 1, further comprising detecting the failure due to a failure to receive the status message responsive to a health check message sent from the load balancer to the first server.

5. The method of claim 1, further comprising determining if the first server is a primary server for the first service, and if not ignoring the failure for the first service and not automatically enabling the second server to execute the first service.

6. The method of claim 1, further comprising generating a failover rule including a plurality of entries each having a service identifier, a primary server identifier and a secondary server identifier, and storing the failover rule in the load balancer.

7. The method of claim 6, further comprising generating a service rule including a plurality of entries each having a service identifier and a primary server identifier, and storing the service rule in the load balancer.

8. The method of claim 1, further comprising:

designating a first port on the first server as a master port, and designating a second port on the first server as a slave port; and

designating a first port on the second server as a first replication port, and designating a second port on the second server as a second replication port.

9. The method of claim 8, further comprising replicating data updates to the first port on the first server to the second port on the first server and to the first and second ports on the second server.

10. The method of claim 8, further comprising responsive to the failure of the first service, designating the first port on the second server as the master port.

11. The method of claim 10, after the failure has been resolved, designating the first port on the first server as the master port.

12. A load balancer comprising:

a failover logic unit to identify a plurality of servers coupled behind the load balancer to execute a plurality of services, generate and store in the load balancer a failover rule including a plurality of entries each having a service identifier, a primary server identifier and a secondary service identifier, generate and store in the load balancer a service rule including a plurality of entries each having a service identifier and a primary server identifier, and to determine a failure in a first server that executes a first service responsive to a lack of response by the first server to a keepalive message sent by the load balancer to the first server.

13. The load balancer of claim 12, wherein the failover logic unit is further to switch the first service from the first server to a second server based on an entry in the failover rule to enable the second server to handle the service, responsive to the lack of response.

14. The load balancer of claim 12, further comprising a monitoring entity to report the failure of the first server to an administrator.

15. The load balancer of claim 12, wherein, if the first server is a secondary server for a second service, the failover logic unit is to ignore the failure in the first server, and not switch the second service from a primary server for the second service.

16. An article comprising a machine-accessible storage medium including instructions that when executed cause a system to:

receive, in a load balancer, a status message from each of a plurality of services operating on a plurality of servers coupled behind the load balancer;

detect a failure of a first service executing on a first server of the plurality of servers based on a status message; and

automatically enable a second server corresponding to a failover server to the first server to execute the first service, via the load balancer.

17. The article of claim 16, further comprising instructions that when executed enable the system to configure the load balancer as a cluster manager.

18. The article of claim 16, further comprising instructions that when executed enable the system to report the failure via an outside monitoring entity.

19. The article of claim 16, further comprising instructions that when executed enable the system to detect the failure due to a failure to receive the status message responsive to a health check message sent from the load balancer to the first server.

20. The article of claim 16, further comprising instructions that when executed enable the system to determine if the first server is a primary server for the first service, and if not ignore the failure for the first service and not automatically enable the second server to execute the first service.