System and method for providing an online software upgrade in load sharing servers

• US 20040253956A1
• Filed: 06/12/2003
• Published: 12/16/2004
• Est. Priority Date: 06/12/2003
• Status: Active Grant

• Current Assignee: Samsung Electronics Co. Ltd.
• Original Assignee: Samsung Electronics Co. Ltd.
• Inventors: Collins, David Allan
• Granted Patent: US 7,356,577 B2
• Time in Patent Office: Days
• US Class Current: 455/445
• CPC Class Codes: H04W 24/02 Arrangements for optimising operational condition

Abstract

A system and method are disclosed for providing an online software upgrade in a switch in a telecommunications system. The switch includes: 1) a main processing unit for executing call process client applications associated with call connections; 2) call application nodes for executing call process server applications; and 3) system management nodes. A configuration management controller associated with a system management node executes an online software upgrade for the software in the switch while the switch is executing the software that is being upgraded. At least one load sharing policy is capable of associating at least one of the call process client applications with at least one of the call process server applications and to throttle new associations involving the call process server applications.

Claims

1. A switch capable of handling call connections between calling devices and called devices, the switch comprising:

a main processing unit capable of executing call process client applications, each of the-call process client applications associated with one of the call connections;

a plurality of call application nodes capable of executing call process server applications;

a plurality of system management nodes, at least one of the system management nodes associated with a configuration management controller, the configuration management controller capable of executing an online software upgrade for software in the switch while the switch is executing the software to handle the call connections; and

at least one load sharing policy capable of associating at least one of the call process client applications with at least one of the call process server applications and to throttle a number of associations involving the call process server applications.

2. The switch as set forth in claim 1, wherein each of the system management nodes further comprises a system request broker capable of enabling each call process server application to register its service within the system management node.

3. The switch as set forth in claim 2, wherein the configuration management controller is capable of partitioning an address space of the system request broker into a first partition and a second partition.

4. The switch as set forth in claim 3, wherein:

at least two of the call process server applications comprise a load sharing group; and

the load sharing group is accessible from both of the partitions.

5. The switch as set forth in claim 4, wherein the system request broker in each of the system management nodes is capable of maintaining a common service definition and a common address for the load sharing group in each of the partitions.

6. The switch as set forth in claim 4, wherein:

the system request broker in each of the system management nodes is associated with one of the partitions; and

the system request broker in one of the system management nodes is capable of;

receiving a request from at least one of the call process client applications;

extracting a partition identifier from the request;

determining whether the extracted partition identifier identifies the partition associated with the system request broker;

handling the request if the extracted partition identifier identifies the partition associated with the system request broker; and

communicating the request to the system request broker in the other system management node if the extracted partition identifier does not identify the partition associated with the system request broker.

7. The switch as set forth in claim 4, wherein the at least one load sharing policy specifies that:

a throttling factor associated with one of the call process server applications is initialized to one hundred percent when the call process server application joins the load sharing group in the first partition; and

the throttling factor is initialized to zero percent when the call process server application joins the load sharing group in the second partition, wherein the throttling factor is increased from zero percent during a test of the call process server application.

8. The switch as set forth in claim 1, wherein the at least one load sharing policy associates one of the call process client applications with one of the call process server applications using a round-robin algorithm, the round-robin algorithm selecting one of the call application nodes.

9. The switch as set forth in claim 1, wherein the at least one load sharing policy comprises:

one or more CPU-based throttling factors used when the online software upgrade is not occurring; and

one or more upgrade throttling factors used when the online software upgrade is occurring.

10. The switch as set forth in claim 9, wherein the one or more CPU-based throttling factors and the one or more upgrade throttling factors are both used when the online software upgrade is occurring.

11. The switch as set forth in claim 1, wherein:

at least two of the call process server applications comprise a load sharing group; and

the load sharing policy is encapsulated in a session object that is capable of connecting to the load sharing group, the session object capable of transporting messages between one of the call process client applications and one of the call process server applications in the load sharing group.

12. The switch as set forth in claim 3, wherein the configuration management controller is capable of:

selecting a backup system management node and at least one new call application node for the second partition and installing new software in the second partition;

assigning a new system request broker address to the second partition that is different than a system request broker address assigned to the first partition;

starting the backup system management node in the second partition with the new system request broker address;

testing the new software after the new software has been loaded into the second partition;

in response to determining that the new software passes the test, shutting down each node in the first partition, installing the new software in the second partition, changing the system request broker address to the second partition, and restarting nodes in the second partition; and

in response to determining that the new software fails the test, shutting down each node in the second partition, using a previous version of the software in the first partition, changing the system request broker address to the first partition, and restarting nodes in the first partition.

13. A wireless network, comprising:

a plurality of base stations capable of communicating with a plurality of mobile stations; and

a mobile switching center coupled to the plurality of base stations, the mobile switching center capable of handling call connections involving at least one of the mobile stations, the mobile switching center comprising;

a main processing unit capable of executing call process client applications, each of the call process client applications associated with one of the call connections;

a plurality of call application nodes capable of executing call process server applications;

a plurality of system management nodes, at least one of the system management nodes associated with a configuration management controller, the configuration management controller capable of executing an online software upgrade for software in the mobile switching center while the mobile switching center is executing the software to handle the call connections; and

at least one load sharing policy capable of associating at least one of the call process client applications with at least one of the call process server applications and to throttle a number of associations involving the call process server applications.

14. The wireless network as set forth in claim 13, wherein:

each of the system management nodes further comprises a system request broker capable of enabling each call process server application to register its service within the system management node; and

the configuration management controller is capable of partitioning an address space of the system request broker into a first partition and a second partition.

15. The wireless network as set forth in claim 14, wherein:

at least two of the call process server applications comprise a load sharing group; and

the load sharing group is accessible from both of the partitions.

16. The wireless network as set forth in claim 15, wherein:

the system request broker in each of the system management nodes is capable of maintaining a common service definition and a common address for the load sharing group in each of the partitions;

the system request broker in each of the system management nodes is associated with one of the partitions; and

the system request broker in one of the system management nodes is capable of;

receiving a request from at least one of the call process client applications;

extracting a partition identifier from the request;

determining whether the extracted partition identifier identifies the partition associated with the system request broker;

handling the request if the extracted partition identifier identifies the partition associated with the system request broker; and

communicating the request to the system request broker in the other system management node if the extracted partition identifier does not identify the partition associated with the system request broker.

17. The wireless network as set forth in claim 15, wherein the at least one load sharing policy comprises:

a round-robin algorithm that associates one of the call process client applications with one of the call process server applications by selecting one of the call application nodes;

a throttling factor associated with one of the call process server applications that is initialized to one hundred percent when the call process server application joins the load sharing group in the first partition; and

the throttling factor that is initialized to zero percent when the call process server application joins the load sharing group in the second partition, wherein the throttling factor is increased from zero percent during a test of the call process server application.

18. The wireless network as set forth in claim 13, wherein the at least one load sharing policy comprises:

one or more CPU-based throttling factors used when the online software upgrade is not occurring; and

one or more upgrade throttling factors used when the online software upgrade is occurring.

19. The wireless network as set forth in claim 18, wherein the one or more CPU-based throttling factors and the one or more upgrade throttling factors are both used when the online software upgrade is occurring.

20. The wireless network as set forth in claim 13, wherein:

at least two of the call process server applications comprise a load sharing group; and

the load sharing policy is encapsulated in a session object that is capable of connecting to the load sharing group, the session object capable of transporting messages between one of the call process client applications and one of the call process server applications in the load sharing group.

21. The wireless network as set forth-in claim 15, wherein the configuration management controller is capable of:

selecting a backup system management node and at least one new call application node for the second partition and installing new software in the second partition;

assigning a new system request broker address to the second partition that is different than a system request broker address assigned to the first partition;

starting the backup system management node in the second partition with the new system request broker address;

testing the new software after the new software has been loaded into the second partition;

in response to determining that the new software passes the test, shutting down each node in the first partition, installing the new software in the second partition, changing the system request broker address to the second partition, and restarting nodes in the second partition; and

in response to determining that the new software fails the test, shutting down each node in the second partition, using a previous version of the software in the first partition, changing the system request broker address to the first partition, and restarting nodes in the first partition.

22. For use in a switch that is capable of handling call connections between calling devices and called devices, the switch comprising:

a main processing unit capable of executing call process client applications, each of the call process client applications associated with one of the call connections;

a plurality of call application nodes capable of executing call process server applications; and

a plurality of system management nodes;

a method for providing an online upgrade for software in the switch, the method comprising;

executing in a configuration management controller an online software upgrade for software in the switch while the switch is executing the software to handle the call connections, the configuration management controller associated with at least one of the plurality of system management nodes; and

enforcing at least one load sharing policy capable of associating at least one of the call process client applications with at least one of the call process server applications and to throttle a number of associations involving the call process server applications.

23. The method as set forth in claim 22, further comprising:

enabling each call process server application to register its service within one of the system management nodes using a system request broker in each of the system management nodes; and

partitioning an address space of the system request broker into a first partition and a second partition.

24. The method as set forth in claim 23, wherein:

at least two of the call process server applications comprise a load sharing group; and

the load sharing group is accessible from both of the partitions.

25. The method as set forth in claim 24, further comprising:

maintaining, in the system request broker in each of the system management nodes, a common service definition and a common address for the load sharing group in each of the partitions, the system request broker in each of the system management nodes associated with one of the partitions;

receiving a request from at least one of the call process client applications at the system request broker in one of the system management nodes;

extracting a partition identifier from the request;

determining whether the extracted partition identifier identifies the partition associated with the system request broker;

handling the request if the extracted partition identifier identifies the partition associated with the system request broker; and

communicating the request to the system request broker in the other system management node if the extracted partition identifier does not identify the partition associated with the system request broker.

26. The method as set forth in claim 24, wherein the at least one load sharing policy comprises:

a round-robin algorithm that associates one of the call process client applications with one of the call process server applications by selecting one of the call application nodes;

a throttling factor associated with one of the call process server applications that is initialized to one hundred percent when the call process server application joins the load sharing group in the first partition; and

the throttling factor that is initialized to zero percent when the call process server application joins the load sharing group in the second partition, wherein the throttling factor is increased from zero percent during a test of the call process server application.

27. The method as set forth in claim 22, wherein the at least one load-sharing policy comprises:

one or more CPU-based throttling factors used when the online software upgrade is not occurring; and

one or more upgrade throttling factors used when the online software upgrade is occurring.

28. The method as set forth in claim 27, wherein the one or more CPU-based throttling factors and the one or more upgrade throttling factors are both used when the online software upgrade is occurring.

29. The method as set forth in claim 22, wherein at least two of the call process server applications comprise a load sharing group;

and

further comprising encapsulating the load sharing policy in a session object that is capable of connecting to the load sharing group, the session object capable of transporting messages between one of the call process client applications and one of the call process server applications in the load sharing group.

30. The method as set forth in claim 24, wherein executing the online software upgrade in the configuration management controller comprises:

selecting a backup system management node and at least one new call application node for the second partition and installing new software in the second partition;

assigning a new system request broker address to the second partition that is different than a system request broker address assigned to the first partition;

starting the backup system management node in the second partition with the new system request broker address;

testing the new software after the new software has been loaded into the second partition;

in response to determining that the new software passes the test, shutting down each node in the first partition, installing the new software in the second partition, changing the system request broker address to the second partition, and restarting nodes in the second partition; and

in response to determining that the new software fails the test, shutting down each node in the second partition, using a previous version-of the software in the first partition, changing the system request broker address to the first partition, and restarting nodes in the first partition.