CN114157618A - Data exchange module and switch supporting FCoE service - Google Patents

Abstract

The invention discloses a data exchange module and a switch supporting FCoE service, wherein the scheme adopts an exchange chip CTC7148 to form a line card main chip and construct a corresponding adaptation layer so as to construct an FCoE functional module on the exchange chip CTC 7148; the adaptation layer is a HAL hardware abstraction layer, the HAL hardware abstraction layer is composed of a software layer and a hardware adaptation layer, the software layer provides a uniform function interface to the outside, and the hardware adaptation layer corresponds to each chip platform. The invention constructs an internet fusion enhanced network (CEE) module for fusing FC switching and Ethernet switching data center network application based on a Shengke switching chip CTC7148, and uses the Chinese Shengke chip CTC7148 as a main chip to realize a software and hardware solution, thereby effectively overcoming the problems in the prior art.

Description

Data exchange module and switch supporting FCoE service

Technical Field

The invention relates to a network equipment design scheme, in particular to an equipment architecture scheme of a data center switch and a core switch.

Background

In the former data center environment network, server devices are connected to a storage network (SAN) through fc (fiber channel) switches, the server devices are connected to an IP network (LAN) through ethernet switches, and the server devices are connected to the IP network and the storage network at the same time. Therefore, two networks exist in the data center at the same time for deployment and operation and maintenance. Meanwhile, the existing FC storage networks are mutually independent closed networks and cannot perform information interaction with each other.

The fcoe (fiber Channel over ethernet) technology converts an FC packet into an IP packet, so that the FC packet can be transmitted in an IP network, and simultaneously, the IP packet can also be converted into a FC packet and transmitted in an FC storage network. Therefore, the original Ethernet switching and FC switching networks can be unified, and communication between two independent FC networks can be realized through the Internet. The FCoE converged network can support LAN and SAN data types, the number of data center equipment and cables is reduced, points needing to be supported are reduced after convergence into a unified network, and management burden is reduced.

The FCoE is combined with virtualization technology to support the virtualization of one device comprising a plurality of devices as a logic device. Multiple links crossing equipment can be bundled into a high-bandwidth link, only the bandwidth is reduced when the link or the equipment fails, the on-off of the line is not influenced, the service is further interrupted, and the reliability and the availability of the network are greatly improved.

The FCoE function is integrated into a large-capacity exchange module, and high-bandwidth data interaction between the SAN storage network and the IP Ethernet is achieved.

However, the existing FCoE device has a single machine form and limited bandwidth capacity, and cannot meet the requirements of practical application.

Disclosure of Invention

Aiming at the problems of the existing equipment architecture scheme of a data center switch and a core switch, the invention aims to provide a data exchange module supporting FCoE services and a switch architecture scheme based on the data exchange module, which can integrate the FCoE services into the exchange module, fully utilize the high bandwidth of the exchange module and flexibly support line cards of various port types. Therefore, the data center network can be conveniently constructed, and the structure of the data center network is more flexible.

In order to achieve the above purpose, the data exchange module supporting the FCoE service provided by the present invention adopts the exchange chip CTC7148 to form a line card main chip, and constructs a corresponding adaptation layer, so as to construct an FCoE function module on the exchange chip CTC 7148; the adaptation layer is a HAL hardware abstraction layer, the HAL hardware abstraction layer is composed of a software layer and a hardware adaptation layer, the software layer provides a uniform function interface to the outside, and the hardware adaptation layer corresponds to each chip platform.

Further, the HAL hardware abstraction layer is provided with L2, L3 and FCoE hardware adaptation interface.

Further, the data exchange module adopts a chip BCM88060 to construct an FC function module.

Further, the exchange chip CTC7148 in the data exchange module is connected with the chip BCM88060 through 4 SerDes to form a 32G FC port; A16G FC port is formed by the exchange chip CTC7148 connected with the chip BCM88060 through 2 SerDes.

Furthermore, an FC protocol stack supporting FC switching, FCoE, and virtualization is constructed in the data switching module.

Further, the data exchange module adopts a distributed architecture, and is divided into:

1) the line card is used for a service port and is connected with the main control card through a plurality of stack interfaces; detecting whether a port link is down through hardware of a chip, and removing a problematic stack interface from a switching matrix by using information stored in a stack header;

2) the backplane switch card is used for exchanging services among the line card forwarding chips and controlling the size of the backplane switch bandwidth by determining the number of the backplane switch chips;

3) the main control card manages and configures the whole machine.

In order to achieve the above object, the switch supporting FCoE services provided in the present invention has the above data switching module disposed therein.

Furthermore, a switching matrix of a crossbar framework is adopted in the switch to support the switching engine with double master controls.

Furthermore, the line card in the switch is connected with the main control switching engine through a plurality of stack interfaces, and the plurality of stack interfaces of each line card form link aggregation to forward the service packet in a load balancing manner.

Further, if a failure of a stack interface in the line card triggers a port to be broken, and when the failure is detected, the port is removed from the link aggregation, and the remaining ports in the link aggregation continue to bear all services.

The invention constructs an internet fusion enhanced network (CEE) module for fusing FC switching and Ethernet switching data center network application based on a Shengke switching chip CTC7148, and uses the Chinese Shengke chip CTC7148 as a main chip to realize a software and hardware solution, thereby effectively overcoming the problems in the prior art.

The data exchange module scheme provided by the invention realizes the functions of FC exchange and FCoE on the Shengke exchange chip CTC7148, constructs a complete FC protocol stack, supports FC exchange and FCoE and supports virtualization. The scheme develops modules which support different port combinations such as 8G/16Gbps/32G FC interfaces and 10G, 40G and 100G Ethernet interfaces supporting FCoE.

The switch scheme provided by the invention uses a CROSSBAR distributed architecture of the main control switch engine and the line card, simultaneously expands the connection quantity of the switch engine and the stack port of the line card, improves the bandwidth from the line card to the switch engine and increases the redundancy reliability of the port of the backboard. The main control can identify different types of board cards, flexible port configuration is carried out, and various port resources can be fully realized on the same switch. The method can greatly facilitate the construction of the data center network, enables the structure of the data center network to be more flexible, is beneficial to the application and development of virtualization, servers and data migration, and lays a solid foundation for meeting the rapid construction requirement of the internet.

The scheme provided by the invention can be embedded into a server cluster, so that the FC switching network and the Ethernet switching network can be accessed independently, and can also be accessed into the FC switching network and the Ethernet switching network simultaneously, and the two networks are organically integrated through an FCoE technology and are applied to a data center network integrating storage switching and Ethernet switching.

The scheme provided by the invention realizes the fusion of FC switching with low delay and high reliability and the most common and high-speed Ethernet, fully utilizes the high bandwidth of the switching module, flexibly supports line cards of various port types, greatly facilitates the construction of a data center network, and enables the structure of the data center network to be more flexible.

Drawings

The invention is further described below in conjunction with the appended drawings and the detailed description.

FIG. 1 is a block diagram of the overall internal architecture interface of the FCoE protocol stack involved in the present invention;

FIG. 2 is an exemplary diagram of functional layers of an FC protocol stack involved in the present invention;

FIG. 3 is an exemplary diagram of the structure of a home server embedded FCoE module involved in the present invention;

FIG. 4 is an exemplary diagram of the structure of a FC module involved in the present invention;

FIG. 5 is a diagram of an example design of a dual protocol data center switch (FCF) supporting FCoE as referred to in the present invention;

FIG. 6 is an exemplary diagram of a data center switch (FSB) fabric framework involved in the present invention;

fig. 7 is a schematic diagram of data traffic forwarding involved in the present invention;

FIG. 8 is a block diagram showing an example of a 10 slot machine frame structure involved in the present invention;

FIG. 9 is a block diagram showing an example of a 6-slot machine frame structure involved in the present invention;

FIG. 10 is an exemplary block diagram of a 10-slot host system architecture involved in the present invention;

FIG. 11 is an exemplary block diagram of a 6-slot host system architecture involved in the present invention;

FIG. 12 is a diagram of an example of a system data exchange path involved in the present invention;

FIG. 13 is an exemplary diagram of a system management channel involved in the present invention;

fig. 14 is a schematic block diagram of a main control module according to the present invention.

Detailed Description

In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further explained below by combining the specific drawings.

The scheme of the invention innovatively adopts a new generation of high-performance switching chips CTC7148 and BCM88060 in the department to construct an FCoE core module, and when the scheme is applied specifically, a BDROS-II multi-protocol operating system can be used in a matched manner, and the operating system supports a multi-core multi-CPU distributed processing architecture and supports main-standby switching and virtualization; the method can be operated on a single machine, and also can be operated on core routing switching equipment of double master control multi-line cards. The system supports a plurality of protocols of two layers and three layers, including routing protocols such as RIP, OSPF, BGP and ISIS, multicast protocols such as PIM-SM and PIM-DM, an MPLS protocol stack, an IPv4/IPv6 protocol stack, network management protocols such as SNMP and RMON, a plurality of tunnel and VPN protocols and the like, and has a perfect system monitoring and log debugging mechanism.

Referring to fig. 1, it shows a frame example diagram of the overall internal structure interface of the FCoE protocol stack in the FCoE core module according to the present invention.

As can be seen from the figure, the BDROS-II software system operating in the FCoE protocol stack generally mainly includes the following functional modules: the system comprises a drive module, a database part (comprising a name distribution database, a switch node, an N-port node, a routing table, a topology, a configuration database and the like), Service modules (comprising Management Service, Internal Link Service, External Link Service, EEnt Service, Directory Service and Distributed Service) supported by External ports, an FSPF module, a routing table updating module, a Management module (comprising a partition configuration module and a routing table configuration module) and the like.

Wherein, the function of main module is:

1) the BSP module is partially arranged on a layer between the hardware of the mainboard and the operating system, so that in order to support the operating system, a function packet for accessing the register of the hardware equipment is provided for a driver program of the upper layer, and the function packet can better run on the hardware mainboard.

2) The driving module is positioned at the bottom layer, mainly performs hardware initialization and work of allocating memory, interruption and the like, and provides support for the API module. Meanwhile, in order to match the switching chip CTC7148, the program of the module is reprogrammed, a pin GPIO1 of the CPU is connected with an interrupt line of the CTC7148, and the CPU triggers interrupt through the GPIO.

3) And the F port module is mainly used for completing the corresponding functions of the F port in the system, and simultaneously the F port also supports the functions of the NPIV.

4) And the E port module is mainly used for completing the corresponding functions of the E port in the system and realizing the cascade connection of the switches.

5) And the exchange module is mainly used for finishing the exchange function.

6) The management module provides a command line management interface for the user, so that the user can check various information of the configuration switch; setting, downloading and checking a routing table; setting a multicast route; switching port rates; setting a monitoring mirror function; setting port partition function, etc. (the FC protocol stack has independent multicast and routing mechanisms).

For example, the scheme of the invention is based on an implemented switching chip hardware platform, wherein a hardware CPU adopts CN7020, the switching chip is a switching chip model number chip such as a domestic department named CTC8096, CTC7148 and the like, and an FC interface chip uses BCM 88060.

On the basis of the exchange chip hardware platform, a corresponding and complete FC protocol stack is constructed, FC exchange and FCoE are supported, and virtualization is supported.

In the FC protocol stack constructed in this example, an FC functional interface is defined in the HAL adaptation layer, and the FC protocol stack configures the switch chip and the FC MAC through the adaptation layer.

Referring to fig. 2, a configuration example of the FC protocol stack constructed in the present scheme is shown.

As can be seen from the figure, the FC protocol stack constructed in the present solution is entirely composed of a series of functional layers.

Wherein the FC-0 layer describes the physical link between two ports, including specifications of the transmission medium, connectors, transmitters, receivers and their respective characteristics;

the FC-1 layer describes an 8B/10B encoding/decoding scheme;

the FC-2 layer is a frame protocol layer and defines rules and mechanisms for transmitting data blocks, including service types, communication models, segment reassembly, error detection and registration/deregistration services required for coordinating communication between ports;

the FC-3 layer provides a set of services which are general to a plurality of N ports on an FC node, and one-to-many communication is realized;

the FC-4 layer defines the mapping of the fibre channel fabric to existing upper layer protocols such as IP, SCSI, etc.

The FC protocol stack constructed in this example adopts a mechanism of once-searching and multiple-forwarding fast packet forwarding. The FC routing is searched once, the CPU software is firstly processed, then the hardware is issued, and then the hardware is directly processed, so that the number of supported FC hosts is large, the occupied CPU is small, and the efficiency is high.

The scheme of the invention forms a data exchange module supporting FCoE service based on the high-performance exchange chip CTC 7148.

The bandwidth of the CTC7148 of the switching chip reaches 640G, various port modes of 10G to 100G are supported, FCoE message forwarding is supported, and FC routing forwarding based on FCID + FID is supported.

Therefore, the scheme constructs an adaptation layer supporting the exchange chip CTC7148 to realize the construction of the FCoE functional module on the exchange chip CTC 7148.

Specifically, in the scheme, a corresponding HAL hardware abstraction layer is designed, and the HAL hardware abstraction layer is formed by matching a software layer and a hardware adaptation layer.

The software layer provides a uniform functional interface to the outside, and the hardware adaptation layer is the specific implementation of each chip platform.

The hardware adaptation layer is compiled into the system in a macro-cutting pre-compiling mode, so that the chip type is identified through HAL initialization, a corresponding hardware adaptation layer interface is hooked to a function pointer of the software layer, and after success, an upper layer protocol can issue various service functions of the chip.

By way of example, the HAL hardware abstraction layer designs hardware adaptation interfaces such as L2, L3, and FCoE to invoke port or chip based methods.

Therefore, the switching chip CTC7148 realizes a set of adaptation layer interfaces according to the standard interface, and the adaptation of the switching chip CTC7148 is supported.

The port generated by the CTC7148 chip in the data exchange module supporting the FCoE service is a standard Ethernet port, supports FCoE messages, and a native FCoE system can be directly realized by depending on an exchange chip; whereas FC ports require external FC MAC device support.

Accordingly, the FC function module is constructed by adopting the chip BCM88060 in the scheme.

Here BCM88060 supports 4G, 8G, 16G, 32G rate FC ports that can be used to connect FC ports of disk arrays or other switches; in addition, the 16G FC port also supports the configuration of a 10G Ethernet port, and the FC or Ethernet port type is flexibly generated.

Specifically, the scheme is based on 3 HSS modules and 24 SerDes lines in an exchange chip CTC7148, and 4 SerDes in the exchange chip CTC7148 are connected with a chip BCM88060 to form a 32G FC port; meanwhile, 2 SerDes in the switching chip CTC7148 are connected with the SerDes of the chip BCM88060 to form a 16G FC port.

By way of example, in some embodiments of the present solution, an 8FC module and an 8TS-4FC module may be formed.

For an 8FC module, an 8-way 8G/16G/32G FC interface SFP + is provided.

Specifically, by exchanging the 16TE output of the chip CTC7148 to the 2-core chip BCM88060, the 8-way 8G/16G/32G FC interface SFP + can be realized through the 2-core chip BCM 88060.

For an 8TS-4FC module, 8 paths of tera light are provided, and meanwhile, 4 paths of FC interfaces SFP + of 8G/16G/32G are also provided.

Specifically, 8 paths of tera light are output through the switching chip CTC7148, 8 paths of TE are output to a 1-chip BCM88060, and 4 paths of 8G/16G/32G FC interface SFP + are realized through the BCM 88060.

In some embodiments of the present invention, multiple types of ethernet cards, 2CF (2 × 100G), 4QS (4 × 40GE), 16TS (16 terabytes), 24GS-8GT-4TS, 24GT-8GS-4TS, 8FC, and 8TS-4FC, may also be formed in the present invention, so as to implement blending FCoE traffic into the switch module, and make full use of the high bandwidth of the switch module, and flexibly support line cards of various port types.

Referring to fig. 3, a diagram of an exemplary structure of a design scheme of a data exchange module supporting FCoE services that can be embedded in a corresponding server is shown.

As can be seen from the figure, in this module, a china department switching chip CTC8096 provides 14 ports 20GE internally for connecting with a server module, and 8 million SFP + fixed interfaces, 2 service slots externally are provided. The service slot can support 2-port 40GE modules, 4-port ten-gigabit modules and 4-port FC port modules. Wherein, 2-port 40GE modules and 4-port ten-gigabit modules are realized by corresponding port connectors.

Referring to fig. 4, an example of the configuration of the FC interface module with which it is mated is shown. The exemplary scheme forms a 4-port FC interface module based on the chip BCM 88060.

Referring to fig. 5, an example of a dual protocol data center switch (FCF) design supporting FCoE is shown.

As can be seen, the switch scheme is formed based on the combination of the Chinese department switch chip CTC8096 and the chip BCM 88060. The 24-port gigabit Ethernet optical port SFP + interface, the 4-port 100G Ethernet QSFP28 interface and the 24-port 8G/16G/32G FC optical port SFP + interface are provided in the switch scheme, and the switch is capable of being plugged with double power supplies, standard matching with an AC power supply, heat dissipation by a fan, 1U height and 19-inch rack-mounted installation.

Meanwhile, the switch provides both an Ethernet interface and an FC interface, both Ethernet switching and FC switching are supported, and conversion is performed between the Ethernet switching and the FC switching through an FCoE technology.

Referring to fig. 6, an example of a home-made data center switch (FSB) design architecture supporting FCoE is shown.

As can be seen, the switch scheme is based on the chinese shengke switch chip CTC 7148. The switch scheme can provide 24-port gigabit Ethernet optical port SFP + interface and 4-port 100G Ethernet optical port QSFP28 interface, and the switch can be plugged with dual power supplies, is provided with an AC power supply in a standard mode, is cooled by a fan, is 1U high and is installed in a 19-inch rack mode.

The switch solution thus constructed can support both ethernet messages and FCoE messages over an ethernet interface.

The CROSSBAR framework is adopted, a switching engine needs a special fabric chip, the switching engine forwards the messages according to a chip and a port in a stack header, and the switching engine does not support L2 and L3 layer services.

The switch scheme switching engine formed in the scheme of the invention is based on a CTC8096 chip and a CTC7148 chip, and can adopt a distributed architecture in specific implementation.

By way of example, the card-insertion modules in the switch scheme formed in the scheme of the present invention are divided into:

1) the line card is used for the service ports and connected with the main control card through a plurality of stack interfaces, and the stack interfaces transmit service messages in a dynamic load balancing mode so as to ensure that the service messages are not blocked. Whether a port link is down is detected through hardware of the chip, and the problem stack interface is removed from the switching matrix by using information stored in a stack head, so that packet loss through the problem link cannot occur when the message is forwarded.

2) The backplane switch card is used for exchanging services among the line card forwarding chips, controlling the size of backplane switch bandwidth by determining the number of backplane switch chips, and finally determining the line speed forwarding performance of a service port of the whole machine.

3) The main control card and the system software mainly run on the main control card to manage and configure the whole machine. And the system stability can be increased by adopting a double-master control mode, the double-master control mode is mutual backup, the master and the slave control modes respectively store a configuration file and send heartbeat messages to each other, the slave control mode can be upgraded to be the master mode in time when the master control mode has the problems of chain breakage and the like, and the operation of the whole machine is not influenced. And because the main control chip also supports the complete two-layer and three-layer switching function and has table entries such as L2, L3, ACL and the like, when the line card does not transmit the related table entries enough, the main control chip can be switched to the main control transmission mode. For example, the multicast message needs to use the multicast table, the multicast copy table, the MAC, the IPMC and other table entries, the table entry capacity on the line card is limited, and we can directly put some functional table entries on the main control card according to the requirement, thereby maximizing the capabilities of the main control and the line card.

Referring to fig. 7, an example of forwarding data traffic by the switch scheme formed in the present invention is shown.

In the switch scheme formed in the scheme of the invention, the forwarding of data service and the message editing are completed on the line card.

When the message enters from the inlet line card, L2/L3 search is carried out, if the destination port is found to be in the local line card, the message is directly forwarded out without passing through a backboard; if the destination port is in another line card or another chip, the backplane switching chip is required to forward the message received by the ingress line card to the egress line card according to the DestChipId information in the CFlexHeader, and all forwarding entries and configurations related to the service are distributed to each line card.

The uplink port is selected through three-level Hierarchical-Channel Link Aggregation, the hash field is selected flexibly, the Channel-level Link Aggregation support enables DLB to select the uplink port according to dynamic load and other mechanisms to ensure that the flow can be exchanged uniformly through the backboard.

The scheme provided by the invention is further illustrated by specific application examples.

In the embodiment, two high-end machine frame type products are formed based on the design scheme of the core switch supporting various FCoE service modules provided by the invention, and the products have sufficient market adaptability and competitiveness.

Referring to fig. 8 and 9, a configuration example of a 10 slot machine and a configuration example of a 6 slot machine formed based on the switch scheme provided by the present invention are shown.

Referring to fig. 8, the 10-slot product in this example adopts a 12U 19-inch chassis design, 8 service slots, 2 master control slots, 2 power slots, a backplane bandwidth of 1.92Tbps, and supports a switching capacity of 1.92Tbps, with 1+1 backup power supply.

Referring to fig. 9, the 6-slot product in this example employs a 9U 19-inch chassis design, 4 service slots, 2 master slots, and 2 power slots. The backplane bandwidth is 960Gbps, supporting 960Gbps of switching capacity. The power supply adopts 1+1 backup.

With further reference to fig. 8 and 9, the overall system of the product of this example includes the following subsystems:

1) the power supply subsystem supplies power to the whole machine and reports the power supply running state to the host subsystem through signals;

2) the heat dissipation subsystem is used for dissipating heat of the whole functional host frame and reporting the running state of the fan to the host frame through signals;

3) the host subsystem is used for completing the necessary functions of all the three-layer switches and providing data exchange services for users, and also completing the monitoring and management of the whole system equipment, comprises a power supply subsystem and a heat dissipation subsystem and is connected to a background network management terminal through a network management interface;

4) and the network management system performs background management on the equipment through 10/100/1000M Ethernet. The host provides an 10/100/1000BASE-TX (100M baseband transmission) optical/electrical combination network port to be connected into the network.

Referring to fig. 10 and fig. 11, the host system in the product provided in this example is composed of 3 kinds of functional modules, namely, a master switching engine (MSU), a service module (LPU, Line Process Unit Line speed processing Unit), and a backplane (BPU, Back Process Unit backplane processing Unit).

The MSU has the main function of providing an operation platform for control software, and has the following main functions:

1) and executing a routing protocol, sending all routing protocol messages to the MSU by the line card for processing, updating routing table entries, and downloading a forwarding table to the line card.

2) And monitoring the running state of the system. The MSU is responsible for collecting operation data of each unit of the system regularly, and generates control information according to operation states of each unit, such as fan acceleration, starting of an alarm box, and the like.

3) As the agent of the network management software, the functions of equipment management, maintenance and the like are provided, and the functions can be performed through the management interfaces (serial ports and network ports) externally provided by the MSU.

4) A data configuration function. The system configuration data, the upgrade software, the system running LOG information and the like are all placed on a control module MSU, and the MSU is provided with FLASH card storage equipment for storing the data files.

Another main function of the MSU is to complete data exchange, and the main functions are: a data exchange platform is provided. And completing a data exchange platform irrelevant to the protocol, wherein after all data messages are processed by the service module and an exchange decision is made, the messages forwarded by the non-service module are exchanged through the exchange engine to complete the data exchange function between the service modules.

The main function of the LPU is to complete the processing and exchange of data, and may complete the following service functions:

1) realizing the linear speed two-layer switching function, and realizing the three-layer switching function according to the issued routing table (forwarding table);

2) realizing the flow classification and flow supervision of the data packet;

3) identifying a control protocol message and sending the control protocol message to a main processing board for processing;

4) the buffer management and scheduling function of the data packet;

5) the back panel BPU provides physical connection between hardware modules, including MSU and LPU modules, and the back panel BPU and corresponding modules are in physical signal connection.

The whole system adopts a Cossbar switching structure, the related data switching channels are shown in figure 12, and the system management channels are shown in figure 13. Each line card is connected to the MSU through a point-to-point high-speed serial differential data bus so as to complete a data exchange function, and the control and management of the main control board on each service module and each exchange engine module are completed in an out-of-band mode (a hundred mega Ethernet).

Referring to fig. 14, the master control module MSU in the product of this example employs the shenke CTC8096 as a switching engine chip. The main control adopts gigabit GE to manage each module, and manages the power supply and the fan through IO (Input/Output) signals of CPLD (Complex Programmable Logic Device). The switching engine chip is the switching core of the scheme, and the switching capacity of the switching chip is 1.2 Tbps. CTC8096 provides 96-way trillion high-speed signals to data switching traffic modules LPUs, which are distributed evenly across the various LPU slots.

The traffic modules in this example product include an FSB module and an FCF module.

The FSB module provides a standard Ethernet port, supports FCoE protocol, and comprises:

1)2CF service module: the port 2 supports the FCoE (FSB)100GE optical interface module, QSFP28 interface. The 100G port may be adapted as a 40GE port;

2)4QS service module: providing 4 paths of FCoE (FSB) -supporting 40GE optical interface modules and QSFP + interfaces;

3)16TS service module: providing 16 paths of tera optical port modules supporting the FCoE (FSB) million self-adaptation, and SFP + interfaces; 4)24GS-8GT-4TS service module: the dual-protocol switching module supports FCoE (FSB), and provides 24-port gigabit Ethernet SFP optical ports, 8-gigabit electric ports and 4-port gigabit SFP + optical ports;

5)24GT-8GS-4TS service module: and the dual-protocol switching module supports FCoE (FSB), and provides a 24-port gigabit Ethernet electrical interface, an 8-port gigabit Ethernet SFP optical interface and a 4-port gigabit SFP + optical interface.

The FCF module directly provides an FC interface, and realizes dual-protocol conversion of FC and FCOE:

1)8FC service module: an 8-port 8G/16G/32G FC optical port SFP + interface is provided;

2)8TS-4FC service module: and 8 paths of tera optical port modules supporting the FCoE (FSB) million self-adaptation, an SFP + interface and a 4-port 8G/16G/32G FC optical port SFP + interface are provided.

The foregoing shows and describes the general principles, essential features, and advantages of the invention. It will be understood by those skilled in the art that the present invention is not limited to the embodiments described above, which are described in the specification and illustrated only to illustrate the principle of the present invention, but that various changes and modifications may be made therein without departing from the spirit and scope of the present invention, which fall within the scope of the invention as claimed. The scope of the invention is defined by the appended claims and equivalents thereof.

Claims (10)

1. The data exchange module supporting the FCoE service is characterized in that an exchange chip CTC7148 is adopted to form a line card main chip, and a corresponding adaptation layer is constructed so as to construct an FCoE functional module on the exchange chip CTC 7148; the adaptation layer is a HAL hardware abstraction layer, the HAL hardware abstraction layer is composed of a software layer and a hardware adaptation layer, the software layer provides a uniform function interface to the outside, and the hardware adaptation layer corresponds to each chip platform.

2. The FCoE service enabled data exchange module of claim 1, wherein the HAL hardware abstraction layer provides L2, L3, and FCoE hardware adaptation interfaces.

3. The FCoE service enabled data exchange module of claim 1, wherein the data exchange module employs the chip BCM88060 to construct the FC functional module.

4. The data exchange module supporting the FCoE service as claimed in claim 3, wherein the CTC7148 of the exchange chip in the data exchange module is connected to the BCM88060 through 4 SerDes to form a 32G FC port; A16G FC port is formed by the exchange chip CTC7148 connected with the chip BCM88060 through 2 SerDes.

5. The FCoE service enabled data exchange module of claim 3, wherein the data exchange module builds a FC protocol stack that supports FC switching, FCoE, and virtualization.

6. The FCoE service capable data exchange module of claim 1, wherein the data exchange module employs a distributed architecture, and comprises:

1) the line card is used for a service port and is connected with the main control card through a plurality of stack interfaces; detecting whether a port link is down through hardware of a chip, and removing a problematic stack interface from a switching matrix by using information stored in a stack header;

2) the backplane switch card is used for exchanging services among the line card forwarding chips and controlling the size of the backplane switch bandwidth by determining the number of the backplane switch chips;

3) the main control card manages and configures the whole machine.

7. Switch supporting FCoE traffic, characterized in that a data switching module according to any of claims 1-6 is arranged in the switch.

8. The switch that supports FCoE services as claimed in claim 7, wherein said switch supports dual-master switching engines using crossbar-fabric switching matrices.

9. The switch supporting FCoE services according to claim 8, wherein the line cards in the switch are connected to the main control switching engine through a plurality of stack interfaces, and the plurality of stack interfaces of each line card form link aggregation to forward the service packets in a load balancing manner.

10. The switch of claim 9, wherein if a failure of a stack interface in the line card triggers a port link drop and the failure is detected, the port is removed from the link aggregation and the remaining ports in the link aggregation continue to carry all traffic.

Images