KR20080070152A - Board redundancy device and method in distributed system - Google Patents

Abstract

The present invention relates to a board redundancy apparatus and method in a distributed system, comprising: a redundancy state management processor for determining a redundancy state (active, standby) of a board and transmitting the redundancy state message; N: M redundancy of the board is possible, including a mapping table management processor that creates a mapping table for connecting virtual slots to real slots using status messages. In addition, the transfer time can be shortened by reflecting the change in the redundancy state of the transferred processing board (New Active) to the mapping table of the line card in the event notification method.

Description

Board redundancy device and method in distributed system {APPARATUS AND METHOD FOR MANAGEING BOARD REDUNDANCY IN DISTRIBUTED SYSTEM}

1 is a simplified block diagram of a distributed system according to an embodiment of the present invention;

2 is a block diagram illustrating a distributed system according to another embodiment of the present invention;

3 is a flowchart illustrating a redundant state management processor in a distributed system according to an embodiment of the present invention;

4 is a flowchart illustrating a mapping table management processor in a distributed system according to an embodiment of the present invention;

5 is a flowchart illustrating redundancy in a distributed system according to an exemplary embodiment of the present invention.

The present invention relates to an apparatus and method for board redundancy in a distributed system, and more particularly, to an apparatus and method for supporting board redundancy using a mapping table connecting virtual slots to actual slots in the distributed system.

Distributed system refers to a system in which various functional boards, such as a processing board and a line card, form a system and share a process. Since the distributed system divides and processes the processor, there are advantages such as resource sharing, load balancing and information sharing, gradual expansion reliability, and availability. In addition, the system is configured to have two or more processing boards having the same configuration and function in preparation for a failure. Many of these processing boards operate as active boards and other predetermined processing boards operate as standbys. Thus, if a failure occurs in the use processing board of the system, the system switches the use processing board to the standby board and also switches the previous standby board to the use processing board.

There are two ways to implement redundancy: the cluster membership model and message duplication. In the cluster membership model, members are joined or separated in a cluster, and an active member is determined through election between members in the cluster, and then the active member is moved to another place in need of the active members. It is a way of notifying the election results. Here, in a distributed system, a group of redundant processing boards becomes a cluster, and each processing board in the redundant group becomes a member.

Another duplication is message duplication. When copying a message to an active member, the message copy is performed by copying the message to transmit the message to all members of the redundant group, and allowing only the active member to receive the message so that the message is delivered to the active member. . However, the message copy is mainly used in the case of 1: 1 duplex board, and may not be implemented depending on the structure of the system. In addition, the traffic in the system may be more than doubled and the load of the standby board may be unnecessarily increased due to the message drop processing on the standby board.

The cluster member model for the dualization implementation, which was introduced at the beginning, does not specifically present a method of informing the active member election and the result of the active member election.

Accordingly, there is a need for a specific apparatus and method for notifying member selection and active member selection results based on a cluster member model.

Accordingly, an object of the present invention is to provide a processing board redundancy apparatus and method in a distributed system.

Another object of the present invention is to provide an N: M board redundancy apparatus and method in a distributed system.

Another object of the present invention is to provide a redundancy apparatus and method using a mapping table that connects a virtual slot and a real slot on which a board is mounted in a distributed system.

According to a first aspect of the present invention for achieving the above objects, a redundancy device of a distributed system, comprising: a redundancy state management processor for determining a redundancy state (active, standby) of a board and transmitting the redundancy state message; And a mapping table management processor for creating a mapping table for connecting a virtual slot to a real slot using a duplication status message.

According to a second aspect of the present invention for achieving the above objects, in the redundancy method of the distributed system, the process of determining the redundancy status (Active, Standby) of the board, the step of transmitting the redundancy status message, And a process of creating a mapping table for connecting virtual slots to real slots using the duplication status message.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. In describing the present invention, when it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, the detailed description thereof will be omitted.

Hereinafter, the present invention will be described a N: M redundancy apparatus and method using a mapping table connecting a virtual slot and a real slot on which a board is mounted in a distributed system. Where N is the number of boards in the active state and M is the number of boards in the standby state.

Figure 1 shows a simplified block diagram of a distributed system according to an embodiment of the present invention.

Referring to FIG. 1, the distributed system includes a line card 100 and processing boards 102, 104, and 106, a redundant state management processor 110 operating within the processing board, and operating within the line card. It consists of a mapping table management processor 112 which creates / updates a virtual slot-real slot mapping table 108.

The line card 100 processes traffic coming into the system from the outside, and the processing boards 102, 104, and 106 receive and process the traffic from the line card 100 in an active state and wait ( In standby, it is used for backup.

The redundancy state management processor 110 determines a redundancy state (Active, Standby) in the processing board (1: 1 redundancy or N: 1 redundancy). For example, the processing boards 102 and 104 mounted in the actual slot 1 and the actual slot 2 in the 2: 1 redundant board distributed system in FIG. 1 are active, and the processing board 106 mounted in the actual slot 3 is active. ) Is determined to be in a standby state. When the redundancy state is determined, the redundancy state process 110 transfers the redundancy state to the mapping table management process 112 of the line card 100.

The mapping table management processor 112 of the line card 100 receiving the redundancy status message sets the contents of the virtual slot-to-real slot mapping table 108 according to the content of the redundancy status message. For example, slot numbers 1 and 2 of the processing boards 102 and 104 determined to be in use are set to actual slot numbers of the mapping table 108. The virtual slot numbers 1 and 2 of the mapping table 108 are set for two processing boards 102 and 104 among the three processing boards 102, 104 and 106 in the distributed system. Here, virtual slot 1 maps to actual slot 1, and virtual slot 2 maps to actual slot 2.

Then, when traffic flows into the line card 100 from the outside (traffic flow), the line card 100 forwards the packet to the processing board 102 according to the contents of the mapping table 108.

2 is a block diagram illustrating a distributed system according to another embodiment of the present invention. Here, the detailed description thereof will be omitted since it is similar to the configuration of the distributed system of FIG. 1.

Referring to FIG. 2, the redundancy state management process 210 monitors whether the redundancy state of the processing boards 202, 204, and 206 is changed, and when the redundancy state change occurs (changeover occurs), the redundancy state management process ( 210 transmits the redundancy state change to the mapping table management process 212 of the line card 200. For example, in the 2: 1 redundant board distribution system of FIG. 1, if the processing board 202 is changed to a standby state and the processing board 204 is changed to a used state, the redundancy state change message is sent to the line card. It passes to the mapping table management process 212 of 200.

When the mapping table management process 212 of the line card 200 receives the redundancy state change message, the mapping table management process 212 updates the contents of the mapping table 208 according to the content of the redundancy state change message. For example, in the mapping table 108 of FIG. 1, the processing board 202 mounted in the actual slot 1 should be changed to the standby state and the processing board 206 mounted in the actual slot 3 should be changed to the use state. Therefore, the actual slot number 1 of the mapping table 208 is changed to 3 and mapped to the virtual slot number 1.

Thereafter, when traffic flows from the outside to the line card 200, the line card 200 is transferred to the processing board 208 according to the updated contents of the mapping table 208.

3 is a flowchart illustrating a redundant state management processor in a distributed system according to an exemplary embodiment of the present invention.

Referring to FIG. 3, the duplication state management process determines the duplication state (Active, Standby) of the processing board (1: 1 duplication or N: 1 duplication) in step 300. The redundancy decision is determined by the user as programmable.

Then, when the redundancy state is determined, the redundancy state process proceeds to step 302 to transfer the redundancy state to the mapping table management process of the line card.

In step 304, the redundancy state management process checks whether the redundancy state (Active, Standby) is changed by monitoring whether the redundancy state is changed in the processing board.

Thereafter, when the redundancy state change occurs, the redundancy state process proceeds to step 306 to transfer the redundancy state change to the mapping table management process of the line card.

The redundant state processor then terminates.

4 is a flowchart illustrating a mapping table management processor in a distributed system according to an exemplary embodiment of the present invention.

Referring to FIG. 4, the mapping table management process of the line card receives a replication status message or a replication status change message from the replication status process in step 400.

In step 402, the mapping table management process updates the contents of the virtual slot-to-real slot mapping table according to the contents of the redundancy status change message. For example, upon receiving an initial 2: 1 redundancy status message, the mapping table management process determines the virtual slot number and the actual slot number of two active processing boards (active) and one standby processing board (standby). Here, it is assumed that the virtual slot number 1 and the actual slot (the slot in which the processing board is actually mounted) are mapped to the number 1, and the virtual slot number 2 and the real slot number 2 are mapped. Thereafter, a problem occurs in the use processing board 1 mounted in the slot 1, and when the use processing board is changed from the active state to the standby state, the existing standby processing board 3 is used. The state changes to the processing board, and the mapping table is updated by changing the actual slot number 1 of the existing processing board 1 to 3.

5 is a flowchart illustrating redundancy in a distributed system according to an exemplary embodiment of the present invention.

Referring to 5, the distributed system determines the redundancy state of the processing board or confirms the redundancy state change of the processing board in step 500.

Thereafter, the distribution system transmits the redundancy state or the redundancy state change to a line card in step 502.

Thereafter, the distributed system updates the virtual slot-to-real slot mapping table information in step 504. By using the virtual slot-to-real slot mapping table, 1: N board redundancy can be implemented. That is, the virtual slot number represents the active number of any processing board in the cluster, and the actual slot refers to the slot number in which the processing board is actually mounted.

In step 506, the distributed system controls the packet processing processor to deliver traffic using the mapping table. For example, when traffic arrives at a line card of a distributed system from outside, packets are forwarded according to the mapping table. If the state of redundancy is changed, the traffic transmitted from the outside to the line card of the distributed system is transferred to the board (if changed from the standby board to the used board) according to the updated mapping table.

The distributed system then terminates the redundancy procedure.

Although the board redundancy of FIG. 1 and FIG. 2 has been described using 2: 1 (in the case of one active board and two standby boards) as an example, N: M board redundancy can be extended.

Meanwhile, in the detailed description of the present invention, specific embodiments have been described, but various modifications are possible without departing from the scope of the present invention. Therefore, the scope of the present invention should not be limited to the described embodiments, but should be determined not only by the scope of the following claims, but also by the equivalents of the claims.

As described above, by using a function that maps virtual slots to actual slots, 1: 1, N: M duplication of processing boards is possible in a distributed system consisting of a line card and a processing board. In addition, the transfer time can be shortened by reflecting the change in the redundancy state of the transferred processing board (New Active) to the mapping table of the line card in the event notification method.

Claims (10)

In the redundancy device of the distributed system, A redundancy state management processor for determining a redundancy state (active, standby) of the board and transmitting the redundancy state message; And a mapping table management processor configured to create a mapping table for connecting a virtual slot and a real slot using the duplication status message. The method of claim 1, And the mapping table includes virtual slot information in which an active board is virtually mounted and actual slot information in which the board is actually mounted. The method of claim 1, And a packet processing processor for forwarding traffic entering the distributed system with reference to the mapping table. The method of claim 1, The redundancy state management process monitors whether the redundancy state of the boards is changed, And when the redundancy state change occurs, forwarding the redundancy state change to a mapping table management process. The method of claim 3, wherein And the mapping table management process updates the contents of the mapping table according to the contents of the duplication state change message when receiving the duplication state change message. In the redundancy method of the distributed system, The process of determining the board's redundancy status (Active, Standby), Transmitting the redundancy status message; And creating a mapping table for connecting a virtual slot and a real slot using the duplication status message. The method of claim 6, The mapping table may include virtual slot information in which an active board is virtually mounted and actual slot information in which the board is actually mounted. The method of claim 6, And forwarding traffic entering the distributed system with reference to the mapping table. The method of claim 6, Monitoring whether the duplication state of the boards is changed; And when the redundancy state change occurs, transferring the redundancy state change to a mapping table management process. The method of claim 9, And updating the contents of the mapping table according to the contents of the duplication state change message when receiving the duplication state change message.

Images