FACILITATING RECOVERY IN A COORDINATED TIMING NETWORK

• US 20080184060A1
• Filed: 10/22/2007
• Published: 07/31/2008
• Est. Priority Date: 01/31/2007
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Dahlen, Dennis J., Crabtree, Donald, Sevigny, Denise M., Carlson, Scott M., Dhondy, Noshir R., Wierbowski, Judith A., Whitney, David E., Elko, David A., Smith, Ronald M.
• Granted Patent: US 7,783,913 B2
• Time in Patent Office: Days
• US Class Current: 714/4
• CPC Class Codes: H04J 3/0641 Change of the master or reference, e.g. take-over or failure of the master

Abstract

Recovery is provided in a timing network. A configuration is defined for that network, and in that configuration, an active primary server is identified that provides a clock source for the network. Additionally, an alternate server is identified that can perform the role of the active primary server, should the active primary server fail. In response to a failure of the primary server, the alternate server detects the failure and performs takeover of the primary server.

Claims

1. An article of manufacture comprising:

at least one computer usable medium having computer readable program code logic to facilitate recovery in a timing network, said computer readable program code logic when executing performing the following;

detecting, by an alternate server of a coordinated timing network having a timing facility for clock synchronization, a failure of a primary active server of the coordinated timing network, said primary active server having a defined role within the coordinated timing network;

taking over by the alternate server the defined role of the primary active server; and

using by the alternate server, in response to taking over, its clock source to synchronize the coordinated timing network.

2. The article of manufacture of claim 1, wherein the detecting comprises one of receiving a system check signal indicating the failure, communicating with a console to learn of the failure, or obtaining an indication of the failure from an arbiter of the coordinated timing network.

3. The article of manufacture of claim 1, wherein the primary active server is the primary active server of a dual server configuration, and wherein the detecting comprises one of receiving a system check signal indicating the failure or communicating with a console to learn of the failure.

4. The article of manufacture of claim 1, wherein the primary active server is the primary active server of a triad server configuration, and wherein the detecting comprises one of obtaining an indication of the failure from an arbiter of the triad server configuration or communicating with a console to learn of the failure.

5. The article of manufacture of claim 1, wherein the primary active server is the primary active server of a dual server configuration, and wherein the taking over comprises:

placing zero or more remaining paths of a path group for the primary active server in an uninitialized state;

setting a stratum level of the alternate server to a defined value indicating it is the primary active server; and

modifying at least one configuration block to indicate the takeover.

6. The article of manufacture of claim 1, wherein the primary active server is the primary active server of a triad server configuration, and wherein the taking over comprises:

placing zero or more remaining paths of a path group for the primary active server in an uninitialized state;

setting a stratum level of the alternate server to a defined value indicating it is the primary active server;

modifying at least one configuration block to indicate the takeover;

putting the alternate server into a takeover mode state; and

issuing to an arbiter server of the triad server configuration a set arbiter takeover mode message.

7. The article of manufacture of claim 1, further comprising surrendering by the primary active server its role as the primary active server, in response to the primary active server detecting it has entered a failure state.

8. The article of manufacture of claim 7, wherein the primary active server is the primary active server of a dual server configuration, and wherein the failure state is in response to entering a check stopped state.

9. The article of manufacture of claim 7, wherein the primary active server is the primary active server of a triad server configuration, and wherein the failure state is entered in response to a loss of attachment to the alternate server and an arbiter server of the triad configuration.

10. The article of manufacture of claim 7, wherein the surrendering comprises:

placing zero or more paths in one or more path groups to an uninitialized state;

setting a stratum level of the primary active server to a given value indicating it is not part of the coordinated timing network;

modifying at least one control block of the surrendering primary active server to indicate the surrender; and

locating a clock source for the surrendering primary active server.

11. The article of manufacture of claim 1, wherein the primary active server is in network communications with the alternate server via a coordinated timing network configured to maintain time synchronization, the primary active server sending server time protocol messages by way of said coordinated timing network.

12. A system to facilitate recovery in a timing network, said system comprising:

an alternate server of a coordinated timing network having a timing facility for clock synchronization to detect a failure of a primary active server of the coordinated timing network, said primary active server having a defined role within the coordinated timing network;

the alternate server to take over the defined role of the primary active server; and

the alternate server, in response to taking over, to use its clock source to synchronize the coordinated timing network.

16. The system of claim 12, wherein the primary active server is the primary active server of a dual server configuration, and wherein the taking over comprises:

placing zero or more remaining paths of a path group for the primary active server in an uninitialized state;

setting a stratum level of the alternate server to a defined value indicating it is the primary active server; and

modifying at least one configuration block to indicate the takeover.

17. The system of claim 12, wherein the primary active server is the primary active server of a triad server configuration, and wherein the taking over comprises:

placing zero or more remaining paths of a path group for the primary active server in an uninitialized state;

setting a stratum level of the alternate server to a defined value indicating it is the primary active server;

modifying at least one configuration block to indicate the takeover;

putting the alternate server into a takeover mode state; and

issuing to an arbiter server of the triad server configuration a set arbiter takeover mode message.

18. The system of claim 12, wherein the primary active server surrenders its role as the primary active server, in response to the primary active server detecting it has entered a failure state.

21. The system of claim 18, wherein the surrendering comprises:

placing zero or more paths in one or more path groups to an uninitialized state;

setting a stratum level of the primary active server to a given value indicating it is not part of the coordinated timing network;

modifying at least one control block of the surrendering primary active server to indicate the surrender; and

locating a clock source for the surrendering primary active server.

22. The system of claim 12, wherein the primary active server is in network communications with the alternate server via a coordinated timing network configured to maintain time synchronization, the primary active server sending server time protocol messages by way of said coordinated timing network.

13. (canceled)

14. (canceled)

15. (canceled)

19. (canceled)

20. (canceled)

23. A method to facilitate recovery in a timing network, said method comprising:

detecting, by an alternate server of a coordinated timing network having a timing facility for clock synchronization, a failure of a primary active server of the coordinated timing network, said primary active server having a defined role within the coordinated timing network;

taking over by the alternate server the defined role of the primary active server; and

using by the alternate server, in response to taking over, its clock source to synchronize the coordinated timing network.

24. The method of claim 23, wherein the primary active server is the primary active server of a dual server configuration, and wherein the taking over comprises:

placing zero or more remaining paths of a path group for the primary active server in an uninitialized state;

setting a stratum level of the alternate server to a defined value indicating it is the primary active server; and

modifying at least one configuration block to indicate the takeover.

25. The method of claim 23, wherein the primary active server is the primary active server of a triad server configuration, and wherein the taking over comprises:

placing zero or more remaining paths of a path group for the primary active server in an uninitialized state;

setting a stratum level of the alternate server to a defined value indicating it is the primary active server;

modifying at least one configuration block to indicate the takeover;

putting the alternate server into a takeover mode state; and

issuing to an arbiter server of the triad server configuration a set arbiter takeover mode message.

26. The method of claim 23, further comprising surrendering by the primary active server its role as the primary active server, in response to the primary active server detecting it has entered a failure state.

27. The method of claim 29, wherein the surrendering comprises:

placing zero or more paths in one or more path groups to an uninitialized state;

setting a stratum level of the primary active server to a given value indicating it is not part of the coordinated timing network;

modifying at least one control block of the surrendering primary active server to indicate the surrender; and

locating a clock source for the surrendering primary active server.