Communication system, communication unit, and communication method

Abstract

We provide a communication system that can more reliably suppress unnecessary communication. [Solution] A system according to one embodiment of the present disclosure is a communication system including redundant first communication unit and second communication unit, wherein the first communication unit and the second communication unit each include a switch unit that performs Layer 2 relay processing according to a learning table and a control unit that manages the data of the learning table in accordance with the following first and second policies. Policy 1: Regarding the addition of entries, one switch unit will add the same entry as the other switch unit has added. Policy 2: Regarding the deletion of entries, only entries whose aging time has expired in both switch sections will be permitted to be deleted.

Description

Technical Field

[0004] , ,

[0005]

[0001] The present disclosure relates to a communication system, a communication unit, and a communication method.

Background Art

[0002] Patent Document 1 describes that in an Ethernet-based (where "Ethernet" is a registered trademark) communication system including two redundant switches, the registration information in the learning table of one switch is transferred to the other switch to synchronize the learning tables of both switches. According to the communication system of Patent Document 1, since the switch does not perform flooding by the above synchronization process, wasteful communication can be suppressed.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] In the synchronization process of transferring the registration information in the learning table to the partner switch as in Patent Document 1, there is a period during which the content of the learning table is asynchronous due to the processing delay associated with the transfer. Therefore, there is a problem that flooding may occur if there is a redundant switchover or the like during such an asynchronous period. In view of such conventional problems, an object of the present disclosure is to provide a communication system that can more reliably suppress wasteful communication.

Means for Solving the Problems

[0005] A system according to one embodiment of the present disclosure is a communication system including redundant first and second communication units, wherein the first and second communication units each include a switch unit that performs Layer 2 relay processing according to a learning table and a control unit that manages the data of the learning table in accordance with the following first and second policies. Policy 1: Regarding the addition of entries, one switch unit will add the same entry as the other switch unit has added. Policy 2: Regarding the deletion of entries, only entries whose aging time has expired in both switch sections will be permitted to be deleted.

[0006] This disclosure can be implemented not only as a system and apparatus having the characteristic configuration described above, but also as a program for causing a computer to execute such characteristic configuration. Furthermore, this disclosure can be implemented as a semiconductor integrated circuit that implements part or all of the system and apparatus. [Effects of the Invention]

[0007] According to this disclosure, unnecessary communications can be suppressed more effectively. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 is a connection diagram showing an example of the overall configuration of a communication network. [Figure 2] Figure 2 is a perspective view showing an example of the external appearance of a communication unit. [Figure 3] Figure 3 is a connection diagram showing an example of a redundant configuration for communication equipment. [Figure 4] Figure 4 shows the state transitions between the active and standby units. [Figure 5] Figure 5 is a block diagram showing an example of the internal configuration of a communication unit. [Figure 6] Figure 6 is a flowchart showing an example of the process for adding an entry. [Figure 7]Figure 7 is a flowchart showing an example of the entry deletion process. [Figure 8] Figure 8 is a connection diagram showing the overall configuration of the communication network in the modified example. [Modes for carrying out the invention]

[0009] <Summary of the embodiments of this disclosure> The embodiments of this disclosure are outlined below.

[0010] (1) A system according to one aspect of this embodiment is a communication system including redundant first communication unit and second communication unit, wherein the first communication unit and the second communication unit each include a switch unit that performs Layer 2 relay processing according to a learning table and a control unit that manages the data of the learning table in accordance with the following first and second policies. Policy 1: Regarding the addition of entries, one switch unit will add the same entry as the other switch unit has added. Policy 2: Regarding the deletion of entries, only entries whose aging time has expired in both switch sections will be permitted to be deleted.

[0011] According to the communication system of this embodiment, since the control unit performs data management in accordance with the first and second policies described above, there is no period in which the learning tables are asynchronous between the switch unit of the first communication unit and the switch unit of the second communication unit. Therefore, flooding can be prevented regardless of whether redundancy switching is present, and wasted communication bandwidth can be more reliably suppressed.

[0012] (2) In the communication system described in (1) above, the control unit may, when an entry is added to the learning table of its own unit, notify the other unit of the data content of the added entry, and when notified by the other unit of the addition of an entry, add the notified entry to the learning table of its own unit. By doing so, it is possible to implement the addition process of entries based on the above-described first policy for the learning table.

[0013] (3) In the communication system of (1) or (2) above, the first communication unit and the second communication unit further have a memory for storing a management table including information indicating whether the aging time has expired for each entry. The control unit prohibits the aging out by the switch unit of its own unit, and when the information has expired in the management tables of both its own unit and the partner unit, it may cause the entry for which the information has expired to be deleted by the switch unit of its own unit. By doing so, it is possible to implement the deletion process of entries based on the above-described second policy for the learning table.

[0014] (4) In the communication system according to any one of (1) to (3) above, when an upper physical link constituting MC-LAG and a lower physical link constituting MC-LAG are connected to the first communication unit and the second communication unit, the first communication unit and the second communication unit may perform load distribution in which both become active. By doing so, it is possible to prevent the occurrence of flooding even when redundant switching is performed for load distribution, and more reliably suppress the waste of the communication bandwidth.

[0015] (5) The method according to one aspect of the present embodiment is a communication method performed in the communication system of (1) to (4) above. Therefore, the communication method of the present embodiment exhibits the same operational effects as the communication systems of (1) to (4) above.

[0016] (6) The apparatus according to one aspect of the present embodiment is a communication unit constituting a sub-combination of the communication systems of (1) to (4) above. Therefore, the communication unit of the present embodiment exhibits the same operational effects as the communication systems of (1) to (4) above.

[0017] <Details of the embodiments of this disclosure> The embodiments of this disclosure will be described in detail below with reference to the drawings. At least some of the embodiments described below may be combined in any way.

[0018] [Overall configuration of the communication network] Figure 1 is a connection diagram showing an example of the overall configuration of the communication network 500. The communication network 500 in this embodiment is a type of Ethernet-based network that uses Ethernet frames for user communication. As shown in Figure 1, the communication network 500 is a communication system that includes a communication device 101 that functions as a trunk-branch line device and a higher-level device 151 that connects to a higher-level network 150.

[0019] The communication device 101 is, for example, an Optical Line Terminal (OLT) on the central office side, and is capable of optical communication with both the upstream trunk optical fiber and the downstream branch optical fiber. The OLT101 consists of two communication units 111 that are capable of redundant switching. Each communication unit 111 works in coordination with each other so that it behaves as if it were a single OLT.

[0020] In this embodiment, of the two communication units 111, one communication unit 111A is the "active system" that is active at startup, and the other communication unit 111B is the "standby system" that is in standby mode at startup. Hereafter, the active communication unit 111A will also be referred to as "active unit 111A," and the standby communication unit 111B will also be referred to as "standby unit 111B."

[0021] Each communication unit 111 has a control card 10 and a line card 20. The control card 10 is a communication node for performing control communication with other communication units 111 and the communication administrator's management device 400 (see Figure 5). The line card 20 is a communication node that performs optical communication with the lower-level device 311 via a branch optical fiber and optical communication with the upper-level device 151 via a trunk optical fiber. In this embodiment, the lower-level device 311 is, for example, an ONU (Optical Network Unit).

[0022] The branch optical fiber 211 is split into two optical fibers 214 and 215 by the optical coupler 213 on the upstream side. Optical fiber 214 is connected to the line card 20 of the active unit 111A, and optical fiber 215 is connected to the line card 20 of the standby unit 111B. The branch optical fiber 211 is branched into multiple optical fibers by the downstream optical coupler 212. A downstream device (ONU) 311, which is the subscriber's optical line termination device, is connected to the downstream end of each optical fiber branched from the downstream optical coupler 212.

[0023] Each communication unit 111's line card 20 is connected to the host device 151 via trunk optical fibers 41 and 42. The line cards 20 of each communication unit 111 are interconnected by physical links 44. The physical links on the uplink side are logically combined into one, forming a multi-chassis link aggregation (MC-LAG). The control cards 10 of each communication unit 111 are interconnected by predetermined control lines 45 (e.g., LAN cables).

[0024] [External structure of the communication unit, etc.] Figure 2 is a perspective view showing an example of the external appearance of the communication unit 111. As shown in Figure 2, the communication unit 111 is a rack-mount type optical communication device employing a slim housing 1. The active unit 111A and standby unit 111B in Figure 1 are composed of the same type of communication unit 111 shown in Figure 2.

[0025] The operating modes that can be set for the communication unit 111 include a "redundant mode" that allows the unit's operating state to be switched between active and standby. Therefore, the communication unit 111 can operate in redundant mode according to the configuration information received from the management device 400.

[0026] The control card 10 is equipped with multiple control ports 11 (three ports 11A, 11B, and 11C in the illustrated example). At least one of the multiple ports 11A, 11B, and 11C is used for transmitting control signals to the control ports of other communication units. The other ports 11A, 11B, and 11C are used for transmitting control signals to the management device 400 and the like.

[0027] The line card 20 includes a downlink port 21 used for PON communication with the ONU 311 and an uplink port 22 used for optical communication with the host device 151. The line card 20 has, for example, 16 downlink ports 21. However, the number of downlink ports 21 included in the line card 20 is not limited as long as it can be accommodated in the enclosure 1.

[0028] The downlink port 21 has a structure that allows for the attachment and detachment of an optical transceiver (not shown) having a predetermined form factor. The downlink ports 21 consist of ports to which, for example, SFP (Small Form-factor Pluggable) transceivers or SFP+ transceivers can be attached and detached. In this case, each downlink port 21 can support a transmission speed of 10Gbps / 1Gbps.

[0029] The uplink port 22 also has a structure that allows for the attachment and detachment of an optical transceiver (not shown) having a predetermined form factor. Uplink port 22 consists of two ports 22A, which can accommodate, for example, QSFP (Quad Small Form Factor Pluggable) 28 transceivers, and four ports 22B, which can accommodate SFP+ transceivers.

[0030] Each of the two ports 22A supports a transmission speed of 100 Gbps, and each of the four ports 22B supports a transmission speed of 10 Gbps. The uplink ports 22 support higher transmission speeds than the downlink ports 21. Thus, since the line card 20 has multiple uplink ports 22, redundancy is possible on the uplink side. Redundancy can be achieved with a single communication unit 111, but in this embodiment, redundancy is achieved using two communication units 111.

[0031] [Configuration of redundant OLT] Figure 3 is a connection diagram showing an example of a redundant configuration of the communication device 101. In Figure 3, "primary" refers to the active unit 111A, which communicates optically with the ONU 311 under normal circumstances. "Secondary" refers to the standby unit 111B, which communicates optically with the ONU 311 in the event of a failure in the active system.

[0032] As shown in Figure 3, in this embodiment, as an example, a 1:1 redundant configuration is adopted for the downlink side of the communication device 101. Specifically, the optical fibers 214 and 215 branching from the optical coupler 213 are optically connected to each communication unit 111A and 111B via optical transceivers mounted on the downlink port 21. The ports 11A of the control cards 10 of each communication unit 111A and 111B are interconnected by control lines 45.

[0033] As mentioned above, MC-LAG is formed by logically bundling the physical uplinks of each communication unit 111A and 111B. Specifically, one uplink port 22A of the active unit 111A and one uplink port 22A of the standby unit 111B are connected to the host device 151 (see Figure 1) via physical links 41 and 42, respectively, and these physical links 41 and 42 are logically combined into one.

[0034] The remaining uplink ports 22A of each communication unit 111A and 111B are used as transit ports. In other words, each unit 111A and 111B are interconnected by a physical link 44 that connects the transit ports. The physical link 44 is implemented using optical fibers and optical transceivers. To implement the physical link 44, either an Active Optical Cable (AOC) or a Direct Attach Cable (DAC) may be used.

[0035] The transmission bandwidth between transit ports is, for example, 100 Gbps. The physical link 44 constitutes the ICL (Inter-Chassis Link). Therefore, under normal circumstances, control traffic and BUM traffic (broadcast, unicast to unknown destinations, and multicast) pass through the ICL.

[0036] [State transitions of the communication unit] Figure 4 is a state transition diagram of the active unit 111A and the standby unit 111B. As shown in Figure 4, when power is supplied to each communication unit 111A and 111B, the line card 20 of the active unit 111A (primary) becomes active (ACT), and the line card 20 of the standby unit 111B (secondary) becomes standby (SBY).

[0037] If the line card 20 of the active unit 111A fails, the control card 10 of the active unit 111A switches its own unit's line card 20 to standby mode and instructs the control card 10 of the standby unit 111B to switch the operating state. The control card 10 of the standby unit 111B activates the operating state of its own unit's line card 20 in response to a switching instruction. In this case, the line card 20 of the standby unit 111B starts communicating with the ONU 311.

[0038] Therefore, the line card 20 of the active standby unit 111B illuminates, while the line card 20 of the active unit 111A, which has transitioned to standby mode, remains in a non-illuminated state. This enables redundant switching. Thus, the redundancy switchover is executed in a relatively short time (so to speak, instantaneously) by a switchover instruction from the active unit 111A to the standby unit 111B.

[0039] When the line card of the active unit 111A is restored, the management device 400 and other devices send a restoration command to the control cards 10 of each unit 111A and 111B. In this case, the control card 10 of the active unit 111A restores the operational state of its own unit's line card 20 to active in response to the recovery command. The control card 10 of the standby unit 111B also restores the operational state of its own unit's line card 20 to standby in response to the recovery command.

[0040] [Internal configuration of the communication unit] Figure 5 is a block diagram showing an example of the internal configuration of communication units 111A and 111B. As shown in Figure 5, the internal configurations of the active unit 111A and the standby unit 111B are the same. Therefore, the internal configuration of the active unit 111A will be described as representative. The active unit 111A includes a control card 10 and a line card 20. The control card 10 is a card-type communication node comprising a control unit 12 and a memory 13.

[0041] The control unit 12 is a arithmetic processing unit including a CPU (Central Processing Unit) that comprehensively controls the operation of the active unit 111A. The control unit 12 may include an FPGA (Field-Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit). The memory 13 is a volatile memory, such as SRAM (Static Random Access Memory) or DRAM (Dynamic Random Access Memory), but may also include a non-volatile memory such as flash memory.

[0042] The control unit 12 can obtain configuration information from the management device 400. The configuration information includes the type of operating mode (redundant mode or standalone mode) and the operating state at startup (active or standby). In the case of the active unit 111A, the control unit 12 activates the unit's operating state at startup. In the case of the standby unit 111B, the control unit 12 sets the unit's operating state to standby at startup.

[0043] The line card 20 is a card-type communication node comprising a signal processing unit 24 and a switch unit 25. The signal processing unit 24 is an LSI (Large Scale Integration) provided in conjunction with the downlink port 21, and functions as an OLT (Output Line Controller) that controls PON communication with the ONU 311 using the optical fiber 214 (or optical fiber 215 in the case of communication unit 111B) and the branch optical fiber 211 as transmission lines.

[0044] The switch unit 25 is an LSI with Layer 2 relay functionality, and is composed of a MAC chip including, for example, at least one of FPGA and ASIC. The switch unit 25 performs relay processing on a per-terminal MAC address basis.

[0045] The uplink frames received by the signal processing unit 24 from the optical fiber 214 (or optical fiber 215 in the case of the standby unit 111B) are forwarded to the host device 151 by the switch unit 25. Downlink frames received by the switch unit 25 from the higher-level device 151 are distributed by the switch unit 25 to the corresponding signal processing unit 24. The signal processing unit 24, upon receiving the downlink frame, sends the received downlink frame to the optical fiber 214 (or optical fiber 215 in the case of the standby unit 111B) in the form of an optical signal.

[0046] The switch unit 25 has a memory area for the MAC learning table (also called the "MAC address table"; hereinafter abbreviated as "learning table") AT. The learning table AT is a table-format database that shows the correspondence between port Pi and destination addresses. "Pi" (i=1,2...) refers to the active physical port of the switch unit Pi, and "i" is the port number.

[0047] In the connection configuration shown in Figure 3, of the multiple physical ports Pi on the switch unit 25, the following ports P1, P2, and P3 are operational. Port P1: Port corresponding to physical link 41 (physical link 42 in the case of standby unit 111B). Port P2: Port corresponding to physical link 45 Port P3: Port corresponding to Downlink Port 21

[0048] If the destination address of the received frame is registered in the learning table AT, the switch unit 25 forwards the received frame to the port Pi corresponding to the destination address. If the destination address of the received frame is not registered in the learning table AT, the switch unit 25 forwards the received frame from all ports Pi except the receiving port (flooding).

[0049] The switch unit 25 performs MAC address learning, i.e., adding entries to the learning table AT, in accordance with the first policy described later. The switch unit 25 deletes entries in the learning table AT that have exceeded a predetermined aging time (age-out).

[0050] [Problems and Solutions of Redundant OLTs] Here, the switch section 25 of the active unit 111A is defined as the "first switch 25A," and the switch section 25 of the standby unit 111B is defined as the "second switch 25B." In the redundant OLT shown in Figure 5, the first switch 25A and the second switch 25B synchronize their learning tables AT with each other in accordance with ICCP (Inter-Chassis Control Protocol) or similar, so that the standby unit 111B can be switched to active at any time.

[0051] However, synchronization requires a certain amount of processing time, so there may be a period during which entries with the same address value (for example, "X") are aged out on the first switch 25A but not on the second switch (hereinafter referred to as the "asynchronous period"). If the standby unit 111B becomes active during this asynchronous period, the first switch 25A may determine that the communication frame with address value X is unlearned and flood the frame, potentially wasting communication bandwidth.

[0052] In this embodiment, in order to resolve the above-mentioned problems, the control cards 10 of the communication units 111A and 111B manage the data of the learning table AT in accordance with a "management policy" which includes the following first and second policies. Furthermore, as a pre-processing step before applying the management policy described below, the control card 10 of the communication units 111A and 111B executes a setting on its own unit's switch section 25A and 25B to prohibit the aging out of the learning table AT.

[0053] (Management policy) Policy 1: When adding entries, both switches 25A and 25B will be synchronized. That is, if one switch unit 25A (or 25B) adds an entry, the other switch unit 25B (or 25A) will also add the same entry. Policy 2: Regarding entry deletion, only entries whose aging time has expired will be permitted to be deleted on both switches 25A and 25B.

[0054] By implementing this data management, asynchronous periods for the learning table AT will not occur between the first switch 25A and the second switch 25B. Therefore, flooding can be prevented regardless of whether redundant switching is performed, and wasted communication bandwidth can be suppressed. To achieve the above data management, a management table MT is constructed in the memory 13 of the control card 10. The management table MT is a table-format data containing entries equivalent to those of the learning table AT, with an additional identification variable RT for each entry indicating whether or not the aging time has expired. RT=0 indicates expiration, and RT≠0 indicates non-expiration (age valid).

[0055] The control unit 12 of the control card 10 monitors the contents of the learning table AT of the switch unit 25, and each time a new entry is generated in the learning table AT, it adds an entry with the same data as the generated entry to the management table MT. Furthermore, with respect to the management table MT, the control unit 12 sets RT=T0 (initial time: for example, 300 seconds) during learning (when an entry is added and when traffic continues), subtracts a predetermined value from RT at predetermined intervals, and clips RT to 0 if it becomes a negative value.

[0056] The predetermined interval mentioned above is measured, for example, by the switch unit 25 autonomously activating a timer. Alternatively, the control unit 12 may manage the timer and instruct the switch unit 25 to subtract (cycle through all entries once) when the timer expires.

[0057] Alternatively, the switch unit 25 may perform a countdown on the learning table AT, and the control unit 12 may copy the age value to the management table MT. This is because counting down all entries would be computationally expensive. Alternatively, instead of copying the age value, non-zero (age valid) and 0 (age expired) values ​​may be recorded. In this case, the switch unit 25 operates in a mode that does not delete entries even if the age value is 0, and can simply relay traffic and initialize the age value when traffic occurs.

[0058] The control unit 12 of the control card 10 of the communication units 111A and 111B uses the management table MT to perform data management of the learning table AT based on the management policy described above. This data management includes "entry addition processing" based on the first policy (Figure 6) and "entry deletion processing" based on the second policy (Figure 7).

[0059] The details of these processes will be explained below with reference to Figures 6 and 7. In the flowcharts of Figures 6 and 7, "own unit" refers to one of the active unit 111A and the standby unit 111B, and "opponent unit" refers to the other of those units 111A and 111B.

[0060] [Adding an entry] Figure 6 is a flowchart showing an example of an entry addition process performed by the control unit 12 of the control card 10 of the communication units 111A and 111B. As shown in Figure 6, the process of adding entries includes "autonomous addition," which is triggered by changes in the learning table AT within the own unit, and "heteronomous addition," which is triggered by notifications from other units.

[0061] Autonomous addition includes a process (step ST11) in which the control unit 12 of the control card 10 monitors whether or not an entry has been added to its own unit's learning table AT. If an entry is added to the learning table AT, the control unit 12 adds the entry added to the learning table AT to its own unit's management table MT (step ST12) and notifies the other unit of the data content of the entry added to the learning table AT (step ST13).

[0062] The externally controlled addition includes a process (step ST21) in which the control unit 12 of the control card 10 monitors whether or not there is a notification of an entry from the other unit. If the control unit receives notification of an entry from the other unit, it adds the notified entry to its own unit's management table MT and learning table AT, respectively (step ST22).

[0063] [Entry Deletion Process] Figure 7 is a flowchart showing an example of an entry deletion process executed by the control unit 12 of the control card 10 of the communication units 111A and 111B. As shown in Figure 7, after a predetermined time (for example, 60 seconds) has elapsed (step ST31), the control unit 12 of the control card 10 generates a management table MT of the other unit through control communication using the control line 45 with the other unit (step ST32). The generated management table MT is temporarily recorded in the memory 13.

[0064] Next, the control unit 12 determines whether or not there is an entry with RT=0 in its own unit's management table MT (step ST33). This determination is made by searching for an entry with RT=0 among the entries contained in its own unit's management table MT. If the result of step ST33 is negative, the control unit 12 terminates the process.

[0065] If the result of step ST33 is positive, the control unit 12 determines whether the same entry that has RT=0 in step ST33 also has RT=0 in the management table MT of the other unit (step ST34).

[0066] If the result of step ST34 is negative, the control unit 12 terminates the process. If the result of step ST34 is positive, the control unit 12 deletes entries where RT=0 for both its own unit and the other unit from its own unit's learning table AT.

[0067] [Examples of communication networks] Figure 8 is a connection diagram showing the overall configuration of the communication network 600 in a modified example. The similarities and differences between the modified communication network 600 (Figure 8) and the above-described communication network 500 (Figure 1) are as follows.

[0068] (match point) 1) The communication device 102 of the communication network 600 is a communication device that includes two redundant communication units 112, 112.

[0069] 2) Communication units 112, 112 have the same device configuration as communication units 111A, 111B (Figures 2 and 5). Therefore, communication unit 112 has the control card 10 and line card 20 shown in Figures 2 and 5. 3) The upper-level physical links 41 and 42 are both connected to the upper-level device 151 and are logically combined into one to form MC-LAG.

[0070] (Differences) 1) The signal processing unit 24 of the line cards 20, 20 does not have PON communication functionality and performs signal control based on a predetermined communication protocol compliant with P2P. Therefore, the communication device 102 of the communication network 600 is not an OLT.

[0071] 2) The physical links (optical fibers 214, 215) connected to the same numbered downlink ports 21 of line cards 20, 20 are all connected to the lower-level device 161 and logically combined into one to form an MC-LAG. 3) Each communication unit 112, 112 is an active unit (primary in Figure 4), and each line card 20, 20 is capable of load balancing, with both being active.

[0072] As shown in the communication network 600 in Figure 8, when each active unit 112,112 performs load balancing, communication frames with the same source address are sent to one of the physical links of the LAG according to a predetermined distribution logic. In this case, a mixture of situations may exist where the aging time has been completed and not completed in one of the active units 112,112 and the other. Therefore, even in the communication network 600 in Figure 8, there is value in managing the learning table AT data in accordance with the management policy described above.

[0073] [Other variations] The embodiments described above are illustrative and not restrictive in all respects. The scope of the present invention is defined by the claims, including the meaning of the scope and equivalents and all modifications within that scope.

[0074] In the above-described embodiment, the control unit 12 and memory 13 may be mounted on the line card 20, and the above-described data management may be performed between the line cards 20 of the communication units 111, 111 (or communication units 112, 112 in the modified example shown in Figure 8). In the embodiments described above, the communication networks 500 and 600 do not necessarily need to use optical fiber as the transmission path for user communication; they may be other Ethernet-based networks, such as a network of L2 switches using LAN cables as the transmission path for user communication. [Explanation of Symbols]

[0075] 1 cabinet 10 Control Cards 11 control ports 11A, 11B, 11C ports 12 Control Unit 13 memory 20 Line Cards 21 Downlink Ports 22 Uplink Ports 22A, 22B Uplink Ports 24 Signal Processing Unit 25 Switch section 25A 1st switch 25B Second switch 41. Optical fiber (physical link) 42. Optical fiber (physical link) 44. Optical fiber (physical link) 45 Control lines 101 Communication equipment (OLT) 102 Communication equipment 111 Communication Unit 112 Communication Unit (Currently in Use) 111A Communication Unit (Currently in use: 1st Communication Unit) 111B Communication Unit (Standby Unit: Second Communication Unit) 150 Top Networks 151 Higher-level equipment 160 Lower Network 161 Subordinate device 200 Management device 211 Branch optical fiber 212 Optical Coupler 213 Optical Coupler 214 Optical Fiber 215 Optical Fiber 311 Subordinate device (ONU) 400 Management device 500 Communication Networks 600 Communication Networks AT Learning Table MT Management Table

Claims

1. A communication system including a redundant first communication unit and a second communication unit, The first communication unit and the second communication unit are A switch unit that performs Layer 2 relay processing according to a learning table, A communication system comprising: a control unit that manages the data of the learning table in accordance with the following first and second policies. Policy 1: Regarding the addition of entries, one switch unit will add the same entry as the other switch unit has added. Policy 2: Regarding the deletion of entries, only entries whose aging time has expired in both switch sections will be permitted to be deleted.

2. The control unit, If an entry is added to the learning table of the unit, the data content of the added entry is notified to the other unit. The communication system according to claim 1, wherein, when notified by the other unit of the addition of an entry, the notified entry is added to the learning table of the own unit.

3. The first communication unit and the second communication unit are It further has memory that stores a management table containing information for each entry indicating whether or not the aging time has expired, The control unit, The switch section of the unit prevents aging out, The communication system according to claim 1 or 2, wherein if the information is full in the management tables of both the own unit and the other unit, the switch unit of the own unit deletes the entry for which the information is full.

4. The first communication unit and the second communication unit are: The upper-level physical links that make up the MC-LAG and the lower-level physical links that make up the MC-LAG are connected. The first communication unit and the second communication unit are A communication system according to claim 1 or claim 2, which performs load balancing in which both parties are active.

5. A communication method performed in a communication system including redundant first and second communication units, The first communication unit and the second communication unit perform relay processing of Layer 2 according to the learning table, A communication method comprising the steps of the first communication unit and the second communication unit managing data in the learning table in accordance with the following first and second policies. Policy 1: Regarding the addition of entries, one switch unit will add the same entry as the other switch unit has added. Policy 2: Regarding the deletion of entries, only entries whose aging time has expired in both switch sections will be permitted to be deleted.

6. A communication unit that is redundant with other communication units, A switch unit that performs Layer 2 relay processing according to a learning table, A communication unit comprising a control unit that manages the data of the learning table in accordance with the following first and second policies. Policy 1: Regarding the addition of entries, the other switch unit or the switch unit of the other unit shall add the same entry that was added by the switch unit of the other unit. Policy 2: Regarding the deletion of entries, only entries whose aging time has expired in both the switch section and the other switch section will be permitted to be deleted.

Images