Interface management architecture

Abstract

The utility model discloses an interface management architecture, which relates to the technical field of computers and comprises an interface management unit and a plurality of controller nodes. The interface management unit comprises an exchange management module and a protocol conversion module, and the exchange management module is used for receiving a control signal from at least one port; and the protocol conversion module is used for determining a target controller node from the plurality of controller nodes according to the control signal, and forwarding the control signal to the target controller node. The architecture solves the technical problem that management interfaces in centralized storage equipment are dispersed.

Description

TECHNICAL FIELD

[0001] The embodiment of the utility model relates to computer field, specifically, relate to an interface management architecture. BACKGROUND

[0002] In prior art, centralized storage device, server, switch and other devices generally have management interface demand, mainstream management interface includes serial port, network port etc. Traditional scheme usually adopts the mode that one controller corresponds to one management interface, that is, each controller is equipped with a serial port and a network port, which are respectively used for local debugging and remote management.

[0003] However, the existing mode management interface is dispersed, and multiple controller nodes cannot be uniformly managed, especially in the multi-controller architecture, the number of management interfaces is large, the wiring is complex, and it is difficult to maintain. That is, the existing management interface has the technical problem of complex layout. UTILITY MODEL CONTENT

[0004] The embodiment of the utility model provides an interface management architecture to at least solve the technical problem of complex layout of the management interface in the prior art.

[0005] According to one embodiment of the utility model, an interface management architecture is provided, comprising an interface management unit and a plurality of controller nodes; the interface management unit comprises a switching management module and a protocol conversion module, the switching management module is used for receiving a control signal from at least one port; the protocol conversion module is used for determining a target controller node from the plurality of controller nodes according to the control signal, and forwarding the control signal to the target controller node.

[0006] Optionally, the protocol conversion module is further used for encapsulating a serial port response signal returned by the target controller node into a network response signal, and the serial port response signal is a response signal matched with the control signal.

[0007] Optionally, each controller node in the plurality of controller nodes comprises a channel control unit, and the channel control unit is used for determining a target channel receiving the control signal from a baseboard management channel and a processor channel according to the control signal.

[0008] Optionally, the at least one port comprises a user port and a technical port, the user port is used for receiving a first control signal in a first working mode, and the technical port is used for receiving a second control signal in a second working mode.

[0009] Optionally, the protocol conversion module is further used for modifying the port number of the port from a first value to a second value in the case of receiving a signal switching message, wherein the first value is used for indicating that the port is a network port, and the second value is used for indicating that the port is a serial port.

[0010] Optionally, the protocol conversion module is further configured to restore the port number from the second value to the first value during the second value is effective, and the preset timer timeout signal is effective.

[0011] The protocol conversion module is further configured to maintain the port number as the second value until the reverse switching message is received, in case that the permanent switching message is detected.

[0012] Optionally, the protocol conversion module is further configured to modify the port number of the port from the third value to the first value in case that the controller node is accessed, wherein the first value is used to indicate that the port is a network port, and the second value is used to indicate that the port is a serial port.

[0013] Optionally, the interface management architecture further comprises at least one reference management unit, wherein each of the at least one reference management unit comprises a reference switching management module and a reference protocol conversion module, and the reference switching management module is configured to receive the control signal from the at least one port.

[0014] The reference protocol conversion module is configured to determine a target controller node from the plurality of controller nodes according to the control signal in case that the interface management unit is in an offline state, and forward the control signal to the target controller node.

[0015] Optionally, the interface management unit and the reference management unit are connected through a backplane heartbeat link, and the interface management unit is determined to be in the offline state in case that three heartbeat frames of the interface management unit are continuously lost.

[0016] Optionally, the interface management architecture further comprises a cascade port, and the cascade port is configured to connect the interface management unit of another machine frame through an Ethernet cable, and the Ethernet cable is used for cascade and sharing between management planes of multiple machine frames.

[0017] The utility model discloses an interface management architecture, which comprises an interface management unit and a reference management unit, wherein the interface management unit comprises a switching management module and a protocol conversion module, the reference management unit is connected with the interface management unit through a backplane heartbeat link, and the switching management module and the protocol conversion module are integrated in the interface management unit. BRIEF DESCRIPTION OF DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the utility model, the following will be briefly introduced the drawings needed to be used in the embodiments, obviously, the drawings in the following description are only some embodiments of the utility model, and for those skilled in the art, other drawings can also be obtained according to these drawings without creative labor.

[0019] Figure 1 It is a schematic diagram of the interface management architecture provided by the embodiments of the utility model.

[0020] Figure 2 A node selection function diagram provided by the embodiment of the utility model for the embodiment of the utility model provides a schematic diagram of path selection function;

[0021] Figure 3 A node selection function diagram provided by the embodiment of the utility model for the embodiment of the utility model provides a schematic diagram of path selection function;

[0022] Figure 4 Another interface management architecture schematic diagram provided by the embodiment of the utility model for the embodiment of the utility model provides another interface management architecture schematic diagram;

[0023] Figure 5 A port number adjustment architecture schematic diagram provided by the embodiment of the utility model for the embodiment of the utility model provides a port number adjustment architecture schematic diagram;

[0024] Figure 6 Another interface management architecture schematic diagram provided by the embodiment of the utility model for the embodiment of the utility model provides another interface management architecture schematic diagram;

[0025] Figure 7 A machine frame architecture schematic diagram provided by the embodiment of the utility model for the embodiment of the utility model provides a machine frame architecture schematic diagram;

[0026] Figure 8 Another interface management architecture schematic diagram provided by the embodiment of the utility model for the embodiment of the utility model provides another interface management architecture schematic diagram. DETAILED DESCRIPTION

[0027] The technical scheme in the embodiments of the utility model will be described clearly and completely below with reference to the drawings in the embodiments of the utility model, apparently, the described embodiments are only part of the embodiments of the utility model, not all the embodiments. Based on the embodiments in the utility model, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the utility model.

[0028] It should be noted that in the description of the utility model, the terms "include", "contain" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or equipment including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or includes the elements inherent to such process, method, article or equipment. The terms "first", "second" and the like in the utility model are used to distinguish similar objects, not to describe a specific order or sequence.

[0029] The embodiments of the utility model provide an interface management architecture, which explains the contents involved in the utility model here:

[0030] 1) BMC: Baseboard Management Controller, baseboard management controller, used for realizing the monitoring, management and remote control function of the server or storage device hardware.

[0031] 2) Ge-Switch: Gigabit Ethernet Switch, used to provide multi-port network data exchange functionality within a device.

[0032] 3) CPU: Central Processing Unit, responsible for executing system instructions, processing data, and controlling the operation of the entire system.

[0033] 4) FPGA: Field Programmable Gate Array, which can be configured with hardware description language to implement flexible data processing and interface control functions.

[0034] 5) CPLD: Complex Programmable Logic Device, commonly used to implement logic control, signal switching and other functions, with high reliability and fast response capability.

[0035] 6) CMC: Customized Management Card, used in centralized storage platforms to achieve unified output and protocol conversion of management interfaces.

[0036] 7) ASIC: Application-Specific Integrated Circuit, a silicon chip customized for specific management functions, used to implement logic control, signal switching and protocol conversion with high integration and low power consumption, and has a faster processing speed and higher reliability than CPLD.

[0037] 8) U-port: User Port, a network interface provided to users for device configuration and management, typically used for routine business configuration.

[0038] 9) T-port: Technology Port, mainly used by professionals for debugging, maintenance and fault diagnosis of equipment.

[0039] According to one aspect of this application, the above interface management architecture is as follows: Figure 1 As shown, it includes: an interface management unit 100 and multiple controller nodes, wherein the multiple controller nodes include a first controller node 102, a second controller node 104, a third controller node 106, and a fourth controller node 108. The interface management unit 100 includes a switch management module 110 and a protocol conversion module 112.

[0040] The exchange management module 110 is configured to receive the control signal from the at least one port 114.

[0041] The protocol conversion module 112 is configured to determine a target controller node from the plurality of controller nodes according to the control signal, and forward the control signal to the target controller node.

[0042] It should be noted that the interface management unit 100 can be a unified access device centrally arranged on a backplane or a separate management card, which provides RJ45 / USB physical ports to the outside, and is interconnected with all controller nodes through a high-speed differential bus. The interface management unit 100 internally integrates at least the exchange management module 110 and the protocol conversion module 112, and is configured to complete the aggregation, exchange, protocol encapsulation / decapsulation, and port mode switching of the multi-controller management message.

[0043] Further, the first controller node 102, the second controller node 104, the third controller node 106, and the fourth controller node 108 can be any control unit with independent computing capability in the chassis, which can be a standard server mainboard, a blade, or a modular storage controller. Each node internally includes a CPU, a BMC, a CPLD, and channel control logic, and is capable of generating or responding to UART / IP messages of the management channel. In the present application Figure 1 , only four are shown, but the number can be expanded to 8, 16, or more according to the backplane switching capability, which still falls within the protection scope of the present application.

[0044] It should be further noted that the exchange management module 110 can be located inside the interface management unit 100 and composed of one or more gigabit or gigabit switching chips, which is responsible for establishing a two-layer or multi-layer forwarding path between the port 114 and each controller node. The exchange management module 110 can support VLAN, QoS, and port mirroring functions, and can dynamically modify the VLAN number under the control of the protocol conversion module 112 to realize instant switching of the network port to the serial port.

[0045] Further, the protocol conversion module 112 can include, but is not limited to, a logic entity implemented in any one or a combination of FPGA, CPLD, and ASIC, which can be configured with the following sub-functions:

[0046] a) parsing the message entering the port 114, and identifying a characteristic string (such as!@# 1 );

[0047] b) rewriting the VLAN number of the corresponding port between 70 to 90 and 80 to 90 according to the identification result;

[0048] c) encapsulating the UART signal on the controller side into a UDP multicast frame, and vice versa.

[0049] d) maintaining a mapping table between the nodes and the channels for determining the target controller node;

[0050] e) sending a takeover trigger signal to the reference management unit upon detecting a primary unit failure.

[0051] The following further describes the port 114.

[0052] In an alternative embodiment, the at least one port 114 can include a user port for receiving the first control signal in the first operating mode and a technology port for receiving the second control signal in the second operating mode.

[0053] It should be noted that the port 114 can be a general term for physical media exposed by the interface management unit 100, including at least two groups of RJ45, user port (U port, User Port) and technology port (T port, Technology Port), and can be extended to USB-Type-C, SFP, etc. The port 114 can be hot switched between the standard network port and the "serial over UDP" port under the cooperation of the switching management module 110 and the protocol conversion module 112, without the need to plug in the cable or restart the device.

[0054] In the above embodiment, the target controller node is not a fixed hardware, which can be used to indicate that the protocol conversion module 112 selects the controller node as the only one that should receive and process the control signal in the current message life cycle. The selection basis can be the node number in the feature string, the VLAN number, the OAM heartbeat state or the explicit instruction of the external management software; after selection, the switching management module 110 immediately establishes a temporary forwarding entry to unicast the downstream traffic to the node, and other nodes are not disturbed.

[0055] Through the above embodiment of the present application, the dispersed multiple management interfaces can be abstracted into a unified interface management unit 100, and high-speed switching is completed inside the interface management unit 100 through the switching management module 110, and the dynamic mapping of the protocol and the port is completed through the protocol conversion module 112, so that the practical new type purposes of one interface managing multiple controls, serial ports being remotely available, and faults being redundant are achieved.

[0056] In an alternative embodiment, the protocol conversion module is further configured to encapsulate a serial port response signal returned by the target controller node into a network response signal, the serial port response signal being a response signal matched with the control signal.

[0057] As Figure 2As shown, in the above embodiment, the protocol conversion module 112 is used to immediately encapsulate the serial port response signal returned by the target controller node (such as the first controller node 102) into a standard network frame, forward it through the SGMII bus, and finally output it to the outside through port 114, thereby completing the transparent bidirectional conversion between serial port and network port.

[0058] Through the above implementation method, the protocol conversion module 112 converts the serial port response signal, which can only be connected locally, into a network response signal that can be accessed remotely in real time, thereby realizing bidirectional interconnection of serial port networking.

[0059] In one alternative implementation, each of the plurality of controller nodes includes a channel control unit, which is used to determine the target channel for receiving control signals from the board management channel and the processor channel based on control signals.

[0060] like Figure 3 As shown, controller node 300 can be connected to interface management unit 100. Controller node 300 can be one of multiple controller nodes in the above embodiments. Its internal structure can include a CPU and a BMC, each corresponding to a serial port channel for transmitting and receiving signals. In the above embodiments, the serial port signals can be processed by the channel control unit in controller node 300, such as the motherboard CPLD. The motherboard CPLD switches the serial port channel by parsing the special identifier input into the CPLD serial port, ensuring that only one path of the serial port in the node is valid externally; that is, for a single node, only the BMC or CPU provides a serial port path externally. Figure 3 The demonstration shows how the motherboard's CPLD switches the serial port to the CPU channel after parsing the serial port data.

[0061] The following combination Figure 4 Another alternative implementation will be described. Figure 4In the figure, the interface management unit 100 collects all management traffic of the first controller node 102 to the fourth controller node 108 through the backplane: the BMC included in the first controller node 102 outputs the first serial port signal, which is sent to the protocol conversion module after being arbitrated by the in-node CPLD, the CPU of the same node provides the second serial port, the third serial port and the fourth serial port signal, all the serial differential pairs are first collected into the switching management module, and then packaged into network frames by the protocol conversion module and assigned with VLAN tags, and finally output through the external port of the switching management module. In the figure, the BMC, the CPU and the associated CPLD of the first controller node 102 constitute a complete substrate management channel and processor channel two-option structure, the remaining second controller node 104, the third controller node 106 and the fourth controller node 108 adopt the same topology, which ensures that the interface management unit 100 only needs one backplane link to switch the target channel between the four nodes, realizes single-point login management of multiple nodes and does not need additional external interfaces.

[0062] In an optional implementation, the protocol conversion module is further configured to modify the port number of the port from a first value to a second value in a case where the signal switching packet is received, wherein the first value is used to indicate that the port is a network port, and the second value is used to indicate that the port is a serial port.

[0063] In an optional implementation, the protocol conversion module is further configured to restore the port number from the second value to the first value in a case where the preset timer timeout signal is effective during the second value is effective.

[0064] The protocol conversion module is further configured to maintain the port number as the second value until the reverse switching packet is received in a case where the permanent switching packet is detected.

[0065] In an optional implementation, the protocol conversion module is further configured to modify the port number of the port from a third value to a first value in a case where the controller node is detected to be accessed, wherein the first value is used to indicate that the port is a network port, and the second value is used to indicate that the port is a serial port.

[0066] The following will be described in combination with Figure 5 The port number adjustment control process of the protocol conversion module is further described.

[0067] As Figure 5As shown, the controller node 506 contains both a BMC 504 and a main CPU (not shown in the figure). Each of them has a serial port (UART) that is converted from serial to parallel by SERDES (SERializer and DESerializer) and then converged to the same protocol conversion module 500. The protocol conversion module 500 encapsulates the two local serial ports into network frames with special identifiers and then sends them back to the switching management module 502 via SERDES, thereby realizing serial-to-network hybrid transmission on the same differential pair.

[0068] Furthermore, when the platform is powered on and all controllers have not yet entered the operating system, the internal algorithm of the protocol conversion module 50 sets the VLAN of the U port and T port to 90 by default. At this time, they appear as two independent "networked serial ports" to the outside world. Remote maintenance personnel can directly interact with the BMC or CPU UART of any node through VLAN 90 packets to complete early debugging, log capture or firmware update.

[0069] Furthermore, when any controller node (such as Figure 5 After the controller node 506 successfully boots into the system, the BMC immediately notifies the protocol conversion module 500 via I2C. The protocol conversion module 500 then invokes a software algorithm to rewrite the U-port VLAN to 70 and the T-port VLAN to 80, restoring them to ordinary Ethernet ports for carrying service management traffic. Simultaneously, the protocol conversion module 500 puts the corresponding SERDES channel into listening mode, continuously scanning for incoming packets.

[0070] If the string "!@#SUART" is found... If the system detects that a user has a temporary serial port requirement, it immediately switches the VLAN of that port back to 90 and starts a 5-minute timer. If there is continuous serial port data within 5 minutes, it keeps VLAN 90; otherwise, it automatically switches back to the original VLAN 70 / 80 and continues to provide network services.

[0071] If the listening string is "!@#SUARTY", it is considered a permanent switch, and the port will continue to work in serial port mode until "!@" is received. The protocol conversion module 500 then controls the BMC again to restore the VLANs of the U port and T port to 70 and 80, and reopen the network channel, after receiving the "net" reset command.

[0072] By means of the above-mentioned embodiments of the present application, the protocol conversion module 500 is combined with the hardware of the exchange management module 502, and the software algorithm of the protocol conversion module 500 is supplemented, so that the present application realizes the multiple functions of “network port-serial port” zero physical switching, remote reversibility and timing automatic recovery on the same SERDES differential pair, guarantees the extreme simplicity of internal wiring of the chassis, and takes into account the flexibility and safety of out-of-band management, thereby successfully solving the problem of shared maintenance interface of multiple controller nodes in a centralized storage platform.

[0073] In an optional embodiment, the interface management architecture further comprises at least one reference management unit, wherein each of the at least one reference management unit comprises a reference exchange management module and a reference protocol conversion module, the reference exchange management module is configured to receive a control signal from at least one port;

[0074] The reference protocol conversion module is configured to determine a target controller node from the multiple controller nodes according to the control signal when the interface management unit is in an offline state, and forward the control signal to the target controller node.

[0075] In an optional embodiment, the interface management unit and the reference management unit are connected through a backplane heartbeat link, and the interface management unit is determined to be in an offline state when three heartbeat frames of the interface management unit are continuously lost.

[0076] It can be understood that in the above-mentioned embodiments, at least one reference management unit can be further included in the interface management architecture, which is used as a redundant design to ensure the stability of the network and improve the high availability and high reliability of the centralized storage device.

[0077] Figure 6 An optional interface management architecture is shown. First, each controller node, including the first controller node 102, the second controller node 104, the third controller node 106 and the fourth controller node 108, internally adopts a double-master serial port composed of a BMC and a CPU. Among them, the first serial port corresponding to the BMC is responsible for out-of-band management, and the serial port corresponding to the CPU includes the second serial port and the fourth serial port, which are used for in-band logging and debugging, and both have the first redundancy in the node.

[0078] Subsequently, the two UARTs are simultaneously connected to the CPLD on the board, and the CPLD can select one as the current effective serial port of the first controller node 102 according to the priority and health degree combination strategy, and the other is in a silent hot standby state, which is the second redundancy.

[0079] When any master (BMC or CPU) in a single node fails, causing its serial port to be unresponsive, the CPLD can automatically switch to the other in milliseconds, thereby ensuring that the node's own management link is not interrupted.

[0080] Further, at the chassis level, the current effective serial ports of the four controller nodes are not wired independently, but are sent to the protocol conversion module 112 for aggregation. In addition, a master-standby structure composed of the interface management unit 100 and the reference management unit 120 is adopted: in the normal state, the interface management unit 100 and the reference management unit 120 simultaneously receive four-node serial port data, but the data is encapsulated into network frames by the switching management module 110 in the interface management unit 100 and sent to the external switch; the standby reference switching management module 124 synchronizes the state in real time but does not send frames externally. Once the main switching management module 110 loses power or the link is abnormal, the reference management unit 120 immediately triggers the switching through the heartbeat loss signal, and transfers the encapsulation and forwarding task to the standby reference conversion module 122 and the reference switching management module 124 within hundreds of milliseconds.

[0081] Through the above-mentioned embodiments of the present application, the countermeasure redundancy design is adopted to ensure the redundancy of the management network, so as to improve the high availability and high reliability of the centralized storage device.

[0082] In an optional embodiment, the above-mentioned interface management architecture further comprises a cascade port, which is used to connect the interface management unit of another chassis through an Ethernet line, and the Ethernet line is used for cascade and sharing between the management planes of multiple chassis.

[0083] As shown in Figure 7 , the first chassis 700 and the second chassis 702 are arranged side by side in the chassis, and the backplane cascade ports of the two chassis are interconnected through a standard Ethernet line, and two sets of RJ45 user ports and technical ports are still exposed externally; the backplane in the chassis aggregates the UART, BMC management SerDes and CPU business SerDes of nodes 0-3 and nodes 4-7 into the FPGA-LFE5U-25F of the respective management card, the FPGA divides the local traffic into VLAN0-99, and carries the cross-chassis frames in VLAN100-199, and the cascade port of any chassis is both the export of external management and the entrance of internal transmission, so that the remote maintenance personnel only need to connect the U / T port at the top of the chassis 700, and can access eight nodes through single sign-on; when any chassis loses power, the FPGA of the other chassis automatically switches the cascade port to the main export after three heartbeat losses, and the MAC and IP of the original U / T port remain unchanged, and the remote session is not aware of the drift, realizing the inter-chassis redundancy; the two chassis only leave two network lines externally, but complete all user, debugging, out-of-band management and serial port networking functions of eight nodes, the wiring is halved, the port is zero-increased, and the investment linearly expands with the nodes without additional switches.

[0084] The following Figure 8 An optional interface management architecture is further described. As shown in Figure 8 , in each single-controller node,Figure 8 Take the controller node 806 as an example, the MAC embedded in the BMC first converts the RGMII into a high-speed SerDes differential signal through the YT8531 PHY (a single-port 10 / 100 / 1000M Ethernet physical layer transceiver), and is directly inserted into the backplane to form a BMC out-of-band management channel; at the same time, the CPU extends two groups of PCIe x1, each of which is hung with an N400 gigabit network card, and each N400 has three SerDes outlets, the first of which goes to the U port, the second of which goes to the T port, and the third of which is left for the internal management frame between the BMC and the CPU. After the three flows are combined, they are also sent into the backplane.

[0085] In addition, in order to achieve complete redundancy, the two groups of SerDes of the BMC and the two groups of N400 SerDes of the CPU are duplicated into A and B two paths, and are respectively dropped on the SW-A of the interface management unit 802 (management card 0) and the SW-B of the interface management unit 804 (management card 1), so that even if one interface management unit or one backplane lane fails, the management traffic can still be out of the frame through another path.

[0086] Inside the interface management unit 802 or the interface management unit 804, four flows can be hard isolated by using VLAN tags. Among them, the VLAN of the U port is 70, the VLAN of the T port is 80, the BMC management VLAN is 40, and the serial port-over-Ethernet VLAN is 90; only two RJ45s-U and T are left outside. When a remote user needs to see the serial port, the FPGA instantly changes the VLAN register of the corresponding port from 70 / 80 to 90, and the same network cable immediately becomes a remote console; after troubleshooting, knock the reverse character, the port tag is instantly restored, and the network session continues-no need to plug and unplug, no need to restart, and two network cables can take away all management functions.

[0087] Through the above-mentioned embodiments of the application, a storage platform management interface network sharing architecture design is provided, aiming to solve the problem of non-aggregation of centralized storage platform management network interfaces, the problem of non-remote debugging of management serial ports, that is, the controller node management interfaces in the centralized storage frame are not interconnected, the management interfaces lack redundant design, and the management interfaces cannot be remotely debugged. The centralized storage frame multi-control device management interface wiring is complex (each controller node has its own network interface and serial port), the management reliability is low (there is no redundant design due to space limitations), and the on-site serial port output log cannot be remotely debugged and viewed (the serial port cannot be remotely output). The application realizes network sharing between the multiple controls of the centralized storage device by adding a network switch chip in the single control node and adding a switch chip on the management card, that is, multi-control sharing of the U interface network and multi-control sharing of the T interface network. The BMC management network is shared between the multiple controllers. The serial port function switching between the multiple controllers is realized by using the existing CPLD and FPGA on the management card in the controller node, and the network of the serial port and the flexible switching of the serial port network output function are realized by using the software algorithm program in the FPGA. The storage platform management interface network sharing architecture design scheme provided by the application can simplify the management interface mode of the centralized storage platform, improve the management efficiency of the centralized storage platform, improve the reliability and availability of the centralized storage platform by completing the symmetrical redundant design, and improve the remote adjustability of the centralized storage platform by increasing the serial port network design function. The scheme improves the management reliability and availability of the storage platform through hardware architecture innovation and software algorithm innovation.

[0088] Through the above description of the embodiments, those skilled in the art can clearly understand that the method according to the above embodiments can be realized by means of software and the necessary general hardware platform, of course, it can also be realized by hardware, but in many cases the former is a better embodiment.

[0089] The interface management architecture provided by the application is described in detail above. The principles and implementation methods of the application are described by applying specific examples in this paper, and the above embodiment description is only applicable to help understand the method and core idea of the application. It should be pointed out that for ordinary skilled persons in the technical field, without departing from the principles of the application, the application can be improved and modified in many ways, and these improvements and modifications also fall within the protection scope of the claims of the application.

Claims

1. An interface management architecture, characterized by, The interface management unit comprises an interface management unit and a plurality of controller nodes; The interface management unit comprises a switching management module and a protocol conversion module, the switching management module is used for receiving a control signal from at least one port; The protocol conversion module is used for determining a target controller node from a plurality of controller nodes according to the control signal, and forwarding the control signal to the target controller node.

2. The interface management architecture of claim 1, wherein, The protocol conversion module is also used for encapsulating a serial port response signal returned by the target controller node into a network response signal, the serial port response signal being a response signal matched with the control signal.

3. The interface management architecture of claim 2, wherein, Each of the plurality of controller nodes comprises a channel control unit, the channel control unit is used for determining a target channel from a substrate management channel and a processor channel according to the control signal, and receiving the control signal.

4. The interface management architecture of claim 2, wherein, At least one of the ports comprises a user port and a technology port, the user port is used for receiving a first control signal in a first working mode, and the technology port is used for receiving a second control signal in a second working mode.

5. The interface management architecture of claim 2, wherein, The protocol conversion module is also used for modifying a port number of the port from a first value to a second value in the case of receiving a signal switching message, wherein the first value is used to indicate that the port is a network port, and the second value is used to indicate that the port is a serial port.

6. The interface management architecture of claim 5, wherein, The protocol conversion module is also used for restoring the port number from the second value to the first value during the validity of the second value and in the case of a preset timer timeout signal being valid; The protocol conversion module is also used for keeping the port number as the second value in the case of detecting a permanent switching message until a reverse switching message is received.

7. The interface management architecture of claim 5, wherein, The protocol conversion module is also used for modifying a port number of the port from a third value to the first value in the case of detecting that the controller node has been accessed.

8. The interface management architecture of claim 1, wherein, At least one reference management unit is further included, wherein each of the at least one reference management unit comprises a reference switching management module and a reference protocol conversion module, the reference switching management module is used for receiving the control signal from at least one of the ports; The reference protocol conversion module is used for determining a target controller node from a plurality of controller nodes according to the control signal and forwarding the control signal to the target controller node in the case that the interface management unit is in an offline state.

9. The interface management architecture of claim 8, wherein, The interface management unit and the reference management unit are connected through a backplane heartbeat link, and the interface management unit is determined to be in an offline state when three heartbeat frames of the interface management unit are continuously lost.

10. The interface management architecture of claim 1, wherein, A cascade port is further included, the cascade port is used for connecting an interface management unit of another machine frame through an Ethernet line, and the Ethernet line is used for cascade and sharing between multi-machine frame management planes.

Images