Selective replay of state information within a computing device

• US 8,014,274 B1
• Filed: 04/16/2008
• Issued: 09/06/2011
• Est. Priority Date: 03/24/2004
• Status: Active Grant

• Current Assignee: Juniper Networks Incorporated
• Original Assignee: Juniper Networks Incorporated
• Inventors: Amicangioli, Anthony D., France, Robert M.
• Primary Examiner(s): Moe; Aung S
• Assistant Examiner(s): RIYAMI, ABDULLA A
• Application Number:
  US12/104,163
• Time in Patent Office: 1,238 Days
• Field of Search: 370/230, 370/216, 370/217, 370/218, 370/219, 370/220, 370/221, 370/228, 370/252, 370/229, 370/352, 370/400, 370/401, 370/236, 370/237, 370/242, 370/241, 370/254, 370/351, 370/353, 370/354, 370/355, 370/356, 370/357, 370/358, 370/389, 370/392, 370/428, 370/429, 714/15, 714/12, 714/6, 714/4, 714/115, 713/100, 713/2, 709/238, 709/240, 709/221, 709/227, 709/228, 709/241, 709/248, 709/204, 709/203, 709/202, 709/201, 709/200, 709/217, 709/218, 709/219, 709/220, 709/222, 709/223, 709/224, 709/225, 709/226, 709/229, 717/131, 717/126, 717/150
• US Class Current: 370/216
• CPC Class Codes: H04L 45/00 Routing or path finding of packets in data switching networks routing or path finding in wireless networks H04W40/00
• CPC Class Codes: H04L 45/586 of virtual routers

Abstract

Techniques are described for synchronizing state information between a plurality of control units. A router, for example, is described that includes a primary control unit and a standby control unit. The primary control unit maintains router resources to ensure operation of the router. To ensure operation, the primary control unit receives state information from the router resources and maintains the state information for consumers, i.e. router resources that require or “consume” state information. The primary control unit performs this state information maintenance process by transmitting update operation messages to consumers and the standby control unit. The consumers respond with an acknowledgement message to both the primary control unit and the standby control unit to inform them that the update has been successfully. The control units use the sequence of these messages to keep all components within the router in sync. Upon assuming control, the standby control unit resumes updating the consumers with state information without having to “relearn” state information, e.g., by way of power cycling the router resources to a known state.

Claims

1. A method comprising:

storing state information within a primary control unit of a network device, wherein different portions of the state information of the primary control unit represent a state of a different one of a plurality of components of the network device;

storing a copy of the state information of the primary control unit within a standby control unit of the network device;

communicating first state data update messages from the primary control unit to the plurality of different components within the network device to update the state of the components, wherein each of the first state data update messages include corresponding state change identifiers;

communicating second state data messages from the primary control unit to the standby control unit to update the state information within the standby control unit, wherein each of the second state data update messages corresponds to one of the first state data update messages and includes the same state change identifier as the corresponding one of the first state data update messages;

upon an occurrence of a failover event from the primary control unit to the standby control unit, determining with the standby control unit which of the portions of the state information stored by the standby control unit is out of synchronization with respect to the state of the components; and

communicating only a subset of the second state data messages from the standby control unit to one or more of the components to reset the state information stored within the components to a condition that matches the state information stored within the standby control unit without sending to the components all of the state information stored in the standby control unit, wherein the subset of second state data messages corresponds to those portions of the state information stored by the standby control unit that are out of synchronization with the state of the corresponding components.

2. The method according to claim 1, wherein communicating only a subset of the second state data messages comprises:

identifying a standby process within the standby control unit containing current state information that is not synchronized state information of a corresponding client process executing within the component; and

transmitting current state information maintained by the standby process to a corresponding client process executing within the component to re-synchronize the state information of the client process and the state information of the standby process.

3. The method according to claim 2,

wherein the state data update messages transmitted by the standby process to the components with the network device are transmitted using different virtual communication channels between the standby process and the corresponding client processes, and

wherein each of the state data update messages includes a process channel identifier that identifies a pair of a primary process and a client process to which the particular state data update message is used to update.

4. The method according to claim 1, wherein the state information includes a current configuration for one or more of the plurality of components.

5. The method according to claim 1, wherein the network device is a router and the state information includes a current state for a forwarding engine that forwards packets.

11. The method of claim 1, further comprising:

receiving acknowledgement messages from the components with the primary control unit, wherein each of the acknowledgement messages corresponds to one of the first state data update messages and includes the same state change identifier as the corresponding one of the first state data update messages; and

forwarding the acknowledgement messages from the primary control unit to the standby control unit.

12. The method according to claim 11, wherein each of the state data update messages and each of the acknowledgement messages comprise a cookie data field permitting the standby process to encode additional data used to reduce the current state information required to re-synchronize the state information of the client process.

6. A network device comprising:

a primary control unit that stores state information for the network device;

a standby control unit for assuming responsibility for managing the state information upon the occurrence of a failover event, the standby control unit storing a copy of the state information; and

a plurality of component that store copies of portions of the state information, the plurality of components communicatively coupled to the primary control unit and the standby control unit;

wherein the primary control unit communicates first state data update messages to the plurality of components to update the copies of the portions of the state information stored within the components, wherein each of the first state data update messages include corresponding state change identifiers,

wherein the primary control unit communicates second state data update messages to the standby control unit of the network device to update the copy of the state information stored within the standby control unit, wherein each of the second state data update messages corresponds to one of the first state data update messages and includes the same state change identifier as the corresponding one of the first state data update messages,

upon an occurrence of a failover event, determining with the standby control unit whether the state information stored by the standby control unit is current with respect to the portions of the state information stored by each of the components; and

when the respective portion of the state information stored by each of the components is not current with respect to the state information of the standby control unit, communicating one or more of the second state data messages from the standby control unit to one or more of the components to reset the portion of the state information stored within the components to a condition that matches the state information stored within the standby control unit.

7. The network device according to claim 6, wherein the standby control unit stores the second state data update messages within a pending message queue.

8. The network device according to claim 6, wherein the state information includes a current configuration for one or more of the plurality of components.

9. The network device according to claim 6, wherein the network device is a router and the state information includes a current state for a forwarding engine.

10. A method comprising:

storing state information within a plurality of components within a network device;

storing state information within a primary control unit, wherein different portions of the state information of the primary control unit represent the state information of the components of the network device;

storing a copy of the state information of the primary control unit within a standby control unit of the network device;

communicating first state data update messages from the primary control unit to the plurality of different components within the network device to update the state information stored within the components, wherein each of the first state data update messages include corresponding state change identifiers;

communicating second state data messages from the primary control unit to the standby control unit to update the state information within the standby control unit, wherein each of the second state data update messages corresponds to one of the first state data update messages and includes the same state change identifier as the corresponding one of the first state data update messages;

upon an occurrence of a failover event, determining with the standby control unit which of the portions of the state information stored by the standby control unit is out of synchronization with respect to the state information stored by the components; and

for those portions of the state information that are out of synchronization, communicating one or more of the second state data messages from the standby control unit to one or more of the components to reset the state information stored within the components to a condition that matches the state information stored within the standby control unit.