WO2016106506A1

WO2016106506A1 - Routing method and device

Info

Publication number: WO2016106506A1
Application number: PCT/CN2014/095351
Authority: WO
Inventors: 胡农达; 冯强
Original assignee: 华为技术有限公司
Priority date: 2014-12-29
Filing date: 2014-12-29
Publication date: 2016-07-07
Also published as: CN106170956B; CN106170956A

Abstract

Provided are a method and device. An operation field can be added to a lookup table entry, and an expression capability of the lookup table entry is enhanced, so as to improve an aggregation capability of the lookup table entry. The method comprises: receiving a data packet, wherein the data packet carries a destination address of the data packet; determining a matching entry corresponding to the destination address in a routing table, wherein an entry of the routing table contains an address operator and a network identifier, and a logical operation result of the operation performed on a significant bit of the network identifier of the matching entry and a significant bit of the destination address according to the address operator of the matching entry is true; and sending the data packet from an output port determined according to output port information in the matching entry. By means of the technical solution disclosed in the present invention, the scale of a lookup table can be reduced, and the demands of a network device for an on-chip storage space are effectively reduced, so as to reduce the area, costs, table lookup time and power consumption of a chip.

Description

Routing method and device

Technical field

The present invention relates to the field of Internet information services, and in particular, to a table lookup routing method and device.

Background technique

Current computer networks, such as Ethernet and the Internet, typically use packet-switched technology (also known as "packet switching") to transmit data. In a packet switching network, when communication is performed between host nodes, communication data is first segmented and encapsulated in a single data packet at the source host node, and injected into the network, and then transmitted and routed through the network device for transmission, and finally the data packet arrives at the destination. The host node is restored and merged into the original communication data. In this process, network devices, such as switches and routers, independently route and forward each data packet, that is, when a data packet arrives at the input port of the network device, the network device based on the address information in the data packet, such as the destination. The MAC address or destination IP address determines the output port of the packet and switches the packet from the input port to the output port. Generally, the process of "determining the output port of a packet" is referred to as "routing," and the process of "switching a packet from an input port to an output port" is referred to as "forwarding." In most current network devices, routing of data packets is performed by looking up a table: one or more logical "lookup tables", also called "routing tables" (such as MAC forwarding tables) or IP routing table), each table contains one or more table entries, each table entry contains information such as address matching information and output port; matching by extracting the address information in the data packet and the address information in the table entry to determine Match the table entries and extract the output port information from the matching table entries to determine the output port. This type of routing is called "table lookup routing."

In computer networks, table lookup routing has a wide range of uses. For example, in an Ethernet built by a Layer 2 switch, a learned forwarding table is stored in the switch, and the table entry includes information such as a MAC address and an output port number. When the switch receives a data packet, it extracts the destination MAC address from the data packet and compares it with the MAC address of the table entry in the forwarding table to determine the matching table entry (the destination MAC address and the MAC address in the table entry). Equal) to get the output port number. In the IP network, the router also performs a similar lookup table routing mechanism: in each router, a lookup table (routing table) learned through a routing protocol is stored, and the table entry of the lookup table includes the network identifier, the network mask, and the output. Information such as port number and next hop network identifier. When the road When the receiver receives a data packet, the destination IP address is extracted from the data packet, and a mask-based matching is performed to determine a matching table entry, thereby obtaining information such as an output port number and a next hop network identifier.

For network devices, the performance of lookup table routing is usually the bottleneck of its performance, and the size of the lookup table is a determining factor in the performance of the table lookup and a key factor in the cost of the network device. Generally, the larger the lookup table, the slower the lookup table, and the more power consumed by the lookup process. At the same time, the larger the lookup table, the more on-chip or off-chip storage is required, which in turn requires the device to have large on-chip or off-chip storage capacity. Because storage capacity, especially on-chip storage capacity, is very expensive, it can significantly increase network equipment costs. More importantly, for a given network device, the on-chip or off-chip storage capacity it has is limited, so the supported table capacity is limited. When the table size exceeds the table capacity, some of the table entries are lost, causing the route to fail. Because the table size is positively related to the size of the network, the size of the table also affects the maximum size of the network that the network device can build. Therefore, new technologies need to be developed to solve this problem.

Summary of the invention

In view of this, the embodiment of the present invention provides a routing method and device, which improves the aggregation capability of a lookup table entry, thereby reducing the size of the lookup table.

In a first aspect, an embodiment of the present invention provides a routing method, including:

Receiving a data packet, where the data packet carries a destination address of the data packet;

Determining, in the routing table, a matching entry corresponding to the destination address, wherein the entry of the routing table includes an address operator and a network identifier, and a valid bit of the network identifier of the matching entry and a valid bit of the destination address The logical operation result of the operation according to the address operator of the matching entry is true;

The data packet is transmitted from an output port determined according to output port information in the matching entry.

With reference to the first aspect, in a first possible implementation, the valid bit of the network identifier is an overall value that is used as the network identifier, and the valid bit of the destination address refers to the destination address. An overall value.

With reference to the first aspect or any of the foregoing possible implementation manners, in a second possible implementation manner, the entry of the routing table further includes a mask, where a valid bit of the destination address is the destination address and the location The mask bit and the result of the operation are an overall value, and the valid bit of the network identifier is an overall value of the network identifier and the result of the mask bit and operation.

In combination with the first aspect or any of the above possible implementations, in a third possible implementation The mask includes a continuous mask and a non-continuous mask.

With reference to the first aspect or any one of the foregoing possible implementation manners, in a fourth possible implementation manner, the entry of the routing table further includes priority information, where the priority information is used to indicate a priority of the belonging entry ;

Determining, in the routing table, a matching entry corresponding to the destination address, comprising: a logic for finding a valid bit of the network identifier of the entry in the routing table and a valid bit of the destination address according to an address operator of the entry The operation result is an entry, and according to the priority information of the found entry, the entry with the highest priority is selected from the searched entry as the matching entry.

With reference to the first aspect or any of the foregoing possible implementation manners, in a fifth possible implementation manner, the determining, in the routing table, a matching entry corresponding to the destination address, including: following a priority from high to low The order in which the logical bit of the network identifier of the entry is found in the routing table and the valid bit of the destination address is logically operated according to the address operator of the entry, and the first entry found is taken as The matching entry, wherein the entries of the lookup table are sorted according to priority.

With reference to the first aspect or any one of the foregoing possible implementation manners, in a sixth possible implementation manner, the sending, by the output port determined according to the output port information in the matching entry, the data packet comprises: according to an address Determining, by the operation result of the output port information in the matching entry, the output port, and transmitting the data packet from the output port, where the address information includes at least one of the following: the destination address, The source address of the data packet, the network identifier of the matching entry, and the current node address.

With reference to the first aspect or any one of the foregoing possible implementation manners, in a seventh possible implementation, the sending, by the output port determined according to the output port information in the matching entry, the data packet includes: according to The hash algorithm selects one of the plurality of ports recorded in the output port information as the output port, and forwards the data packet from the output port.

With reference to the first aspect or any of the foregoing possible implementations, in an eighth possible implementation, the matching entry further includes operation information, where the operation information is used to indicate that the data packet is operated, according to Before the output port determined by the output port information in the matching entry forwards and sends the data packet, the method further includes: operating the data packet according to the operation information in the matching entry.

In combination with the first aspect or any of the possible implementations, the ninth possible implementation The address operator includes: not equal to, greater than, less than, greater than or equal to, or less than or equal to.

In a second aspect, an embodiment of the present invention provides a routing method, including:

Determining a matching entry corresponding to the destination address in the routing table;

And determining, according to the result that the address information is operated according to the output port information in the matching entry, the output port, and sending the data packet from the output port, where the address information includes at least one of the following: The address, the source address of the data packet, the network identifier of the matching entry, and the current node address.

With reference to the second aspect, in a first possible implementation, an entry of the routing table includes an address operator and a network identifier, where a valid bit of the network identifier of the matching entry and a valid bit of the destination address are in accordance with The result of the logical operation of the address operator of the matching entry is true.

With the second aspect or any of the foregoing possible implementation manners, in a second possible implementation manner, the valid bit of the network identifier is an overall value that is used as the network identifier, and a valid bit of the destination address It refers to the overall value of the destination address.

With the second aspect or any of the foregoing possible implementations, in a third possible implementation, the entry of the routing table further includes a mask, where a valid bit of the destination address is the destination address and location The mask bit and the result of the operation are an overall value, and the valid bit of the network identifier is an overall value of the network identifier and the result of the mask bit and operation.

With reference to the second aspect or any of the foregoing possible implementation manners, in a fourth possible implementation manner, the mask includes a continuous mask and a non-continuous mask.

With reference to the second aspect or any of the foregoing possible implementation manners, in a fifth possible implementation manner, the entry of the routing table further includes priority information, where the priority information is used to indicate a priority of the belonging entry ;

With reference to the second aspect or any of the foregoing possible implementation manners, in a sixth possible implementation manner, the determining, in the routing table, a matching entry corresponding to the destination address, including: following a priority from high to low The order of finding the valid bit of the network identifier of the entry in the routing table and the destination address The valid bit is the entry of the logical operation whose operation is performed according to the address operator of the entry, and the first entry found is used as the matching entry, wherein the entries of the lookup table are sorted according to the priority.

With reference to the second aspect or any of the foregoing possible implementation manners, in a seventh possible implementation manner, the matching entry further includes operation information, where the operation information is used to indicate that the data packet is operated, according to Before the output port determined by the output port information in the matching entry forwards and sends the data packet, the method further includes: operating the data packet according to the operation information in the matching entry.

With reference to the second aspect or any of the foregoing possible implementation manners, in an eighth possible implementation manner, the address operator includes: not equal to, greater than, less than, greater than or equal to, or less than or equal to.

In a third aspect, an embodiment of the present invention provides a computing device for routing, including: a processor, a memory, a bus, and a communication interface;

The memory is for storing a computing device execution instruction, the processor is coupled to the memory via the bus, and when the computing device is running, the processor executes the computer-executed instruction stored by the memory to A method of causing the computing device to perform the first aspect or any of the possible implementations of the first aspect, or the method of performing the second aspect or any of the possible implementations of the second aspect.

In a fourth aspect, an embodiment of the present invention provides a routing device, including:

a receiving unit, configured to receive a data packet, where the data packet carries a destination address of the data packet;

a matching unit, configured to determine, in the routing table, a matching entry corresponding to the destination address, where the entry of the routing table includes an address operator and a network identifier, and a valid bit of the network identifier of the matching entry is The logical operation result of the valid bit of the destination address being operated according to the address operator of the matching entry is true;

And a sending unit, configured to send the data packet from an output port determined according to output port information in the matching entry.

With reference to the fourth aspect, in a first possible implementation, the valid bit of the network identifier is an overall value that is used as the network identifier, and the valid bit of the destination address refers to the destination address. An overall value.

With reference to the fourth aspect or any of the foregoing possible implementation manners, in a second possible implementation manner, the entry of the routing table further includes a mask, where a valid bit of the destination address is the destination address and location The mask bit and the result of the operation as an overall value, the valid bit of the network identifier An overall value for the network identification and the result of the mask bit and operation.

With reference to the fourth aspect or any of the foregoing possible implementation manners, in a third possible implementation manner, the mask includes a continuous mask and a non-continuous mask.

With the fourth aspect or any of the foregoing possible implementation manners, in a fourth possible implementation manner, the entry of the routing table further includes priority information, where the priority information is used to indicate a priority of the belonging entry ;

The matching unit is configured to determine, in the routing table, a matching entry corresponding to the destination address, including: the matching unit searches for a valid bit of the network identifier of the entry in the routing table, and a valid bit of the destination address according to the entry The logical operation result of the operation performed by the address operator is a true entry, and based on the priority information of the found entry, an entry having the highest priority is selected from the searched entry as the matching entry.

With reference to the fourth aspect or any one of the foregoing possible implementation manners, in a fifth possible implementation, the matching unit is configured to determine, in the routing table, a matching entry corresponding to the destination address, including: the matching The unit searches for the valid bit of the network identifier of the entry in the order of priority from the highest priority in the order of priority, and the logical bit of the valid address of the destination address is calculated according to the address operator of the entry, and the logical operation result is true. The first entry found is used as the matching entry, wherein the entries of the lookup table are sorted according to priority.

With reference to the fourth aspect or any one of the foregoing possible implementation manners, in a sixth possible implementation, the sending unit is configured to send the data packet from an output port determined according to output port information in the matching entry The method includes: the sending unit determines, according to the result that the address information is operated according to the output port information in the matching entry, the output port, and sends the data packet from the output port, where the address information includes at least the following A: the destination address, a source address of the data packet, a network identifier of the matching entry, and a current node address.

With reference to the fourth aspect or any one of the foregoing possible implementation manners, in a seventh possible implementation, the sending unit is configured to send the data packet from an output port determined according to output port information in the matching entry The method includes: the sending unit selects one port from the plurality of ports recorded in the output port information as the output port according to a hash algorithm, and forwards the data packet from the output port.

With reference to the fourth aspect or any one of the foregoing possible implementation manners, in an eighth possible implementation manner, the device further includes an operation unit, the matching entry further includes operation information, where the operation information is used to indicate a The data packet is operated, and the operating unit is configured to be used according to the matching entry Operation information, operating on the data packet.

With reference to the fourth aspect or any of the foregoing possible implementation manners, in the ninth possible implementation manner, the address operator includes: not equal to, greater than, less than, greater than or equal to, or less than or equal to.

In a fifth aspect, an embodiment of the present invention provides a routing device, including:

a matching unit, configured to determine, in the routing table, a matching entry corresponding to the destination address;

a sending unit, configured to determine, according to the result that the address information is operated according to the output port information in the matching entry, the output port, and send the data packet from the output port, where the address information includes at least one of the following The destination address, the source address of the data packet, the network identifier of the matching entry, and the current node address.

With reference to the fifth aspect, in a first possible implementation, an entry of the routing table includes an address operator and a network identifier, where a valid bit of the network identifier of the matching entry and a valid bit of the destination address are in accordance with The result of the logical operation of the address operator of the matching entry is true.

With reference to the fifth aspect or any of the foregoing possible implementation manners, in a second possible implementation manner, the valid bit of the network identifier is an overall value that is used as the network identifier, and a valid bit of the destination address It refers to the overall value of the destination address.

With reference to the fifth aspect or any one of the foregoing possible implementation manners, in a third possible implementation manner, the entry of the routing table further includes a mask, where a valid bit of the destination address is the destination address and the location The mask bit and the result of the operation are an overall value, and the valid bit of the network identifier is an overall value of the network identifier and the result of the mask bit and operation.

With reference to the fifth aspect or any of the foregoing possible implementation manners, in a fourth possible implementation manner, the mask includes a continuous mask and a non-continuous mask.

With the fifth aspect or any of the foregoing possible implementation manners, in a fifth possible implementation manner, the entry of the routing table further includes priority information, where the priority information is used to indicate a priority of the belonging entry ;

In combination with the fifth aspect or any of the possible implementations, in a sixth possible implementation The matching unit is configured to determine, in the routing table, a matching entry corresponding to the destination address, where the matching unit searches for a valid bit of the network identifier of the entry in the routing table according to a priority from highest to lowest priority. An entry that is a logical operation that operates with the valid bit of the destination address in accordance with the address operator of the entry, and the first entry found as the matching entry, wherein the entry of the lookup table is Priority is sorted.

With reference to the fifth aspect or any one of the foregoing possible implementation manners, in a seventh possible implementation manner, the device further includes an operation unit, where the matching item further includes operation information, where the operation information is used to indicate a The data packet is operated, and the operation unit is configured to operate the data packet according to the operation information in the matching entry.

With reference to the fifth aspect or any of the foregoing possible implementation manners, in the eighth possible implementation manner, the address operator includes: equal to, not equal to, greater than, less than, greater than or equal to, or less than or equal to.

Optionally, in any of the foregoing possible implementation manners, the network identifier includes, but is not limited to, a MAC address, an IP address, and the like.

Optionally, in any one of the foregoing aspects or any of the possible implementations, the comparison items on both sides of the OP are compared as a whole value.

Optionally, in any one of the foregoing aspects or any of the possible implementations, the comparison items on both sides of the OP may be in the order of the first byte to the nth dimension from the high byte to the low byte or from the first dimension. Arrange from the low byte to the high byte order in the nth dimension.

Optionally, in any of the foregoing possible implementation manners, the comparison items on both sides of the OP may further select an m-dimensional label from the n-dimensional labels as an overall value, where the m-dimension The labels are arranged from the 1st dimension to the mth dimension in the order of the high byte to the low byte or from the 1st dimension to the mth dimension in the low byte to the high byte order.

Optionally, in any one of the foregoing possible implementation manners, the priority may be a value that is used to identify the priority of the lookup table entry, and the greater the value, the higher the priority. It can be that the smaller the value, the higher the priority. The priority can also be other expressions that can distinguish between size or order, such as English characters.

Optionally, in any one of the foregoing aspects or any of the possible implementations, the entries in the lookup table are arranged in descending order of priority or low to high.

Optionally, the output port algorithm supports at least one of all mathematical operators such as addition, subtraction, multiplication, division, and shift.

According to the technical solution provided by the embodiment, by adding an operation field to the lookup table entry, the expression capability of the lookup table entry is enhanced, and the aggregation capability of the lookup table entry is improved. Thereby reducing the size of the lookup table, reducing on-chip storage requirements, reducing chip area, cost, table lookup time and power consumption, and improving network scale scalability.

DRAWINGS

In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly described below. It is obvious that the drawings in the following description are only some embodiments of the present invention, Those skilled in the art can also obtain other drawings based on these drawings without paying any creative work.

1 is a schematic diagram of a system logical structure of a routing method application scenario;

2 is a schematic diagram of a system logical structure of a routing method application scenario;

3 is an exemplary flowchart of a routing method according to an embodiment of the invention;

4 is an exemplary flowchart of a routing method according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a lookup table according to an embodiment of the invention; FIG.

6 is a schematic diagram of a system logical structure of a routing method application scenario according to an embodiment of the invention;

FIG. 7 is a schematic diagram of a system logical structure of a routing method application scenario according to an embodiment of the invention; FIG.

FIG. 8 is a schematic diagram of a system logical structure of a routing method application scenario according to an embodiment of the invention; FIG.

FIG. 9 is a schematic diagram showing the logical structure of a routing device according to an embodiment of the invention; FIG.

FIG. 10 is a schematic diagram showing the logical structure of a routing device according to an embodiment of the invention; FIG.

FIG. 11 is a schematic structural diagram of a hardware of a routing system according to an embodiment of the invention.

detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are a part of the embodiments of the present invention, but not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

FIG. 1 is a schematic diagram of a system logical structure of a routing method application scenario according to the present invention. As shown in FIG. 1, the network device 2 forwards the data packet by using the longest prefix matching technology. The technology implements compression of the lookup table by aggregating lookup table entries having the same network identifier prefix and output, and after receiving the data packet, The longest prefix matching technique implements the lookup of matching entries.

Optionally, the network device 2 includes, but is not limited to, a router, a switch, and the like. The network identifier includes, but is not limited to, an Internet Protocol (IP) address, and a Media Access Control (MAC). Address, etc.

The general structure of the technique using a lookup table is shown in Figure 1. Each table entry of the lookup table includes a matching domain and an action domain. The matching domain includes the network identifier, or contains the network identifier and mask, which are used to implement matching of entries when looking up the table. The network identifier may be any address label, and the general representation may be A1.A2.....An; the mask is a prefix continuous mask. The action field contains packet operation information, and generally includes at least output port information. In order to reduce the lookup table size, a network address aggregation based compression technique is used - the lookup table entries with the same network identification prefix and output are combined into one entry to compress the lookup table. As shown in FIG. 1, by aggregating table entries having the same network identifier prefix and output, the six table entries of the original lookup table of the network device 2 can be compressed into four, thereby achieving the purpose of the lookup table compression.

As shown in FIG. 1 , after receiving the data packet sent by the network C, the network device 2 extracts the IP address information of the destination node of the data packet from the data packet, and the destination IP address is 100.50.10.6, according to the compressed lookup table, based on the matching. The condition "(destination address & mask) == (network identification & mask)" performs route lookup, and looks up the entry matching the destination IP address from the compressed lookup table. If there is a matching table entry, the entry with the longest network identification prefix is selected from all matching table entries as the final matching entry, and if there is no matching entry, the default operation set by the network device is performed. As shown in Figure 1, the 3rd and 4th table entries of the lookup table match the IP address. Based on the longest prefix matching principle, the 4th table entry is selected as the final matching entry, and the output port is extracted from the final matching table entry. The information is such that information is provided for packet processing (eg, forwarding). In FIG. 1, the network device 2 forwards the data packet received from the network C from the port 3, thereby implementing routing of the data packet.

The lookup table routing technology based on the longest prefix matching in Figure 1 is efficient for the execution of routing algorithms for simple hierarchical aggregation networks, but the execution of routing algorithms for many high performance networks is inefficient and cannot be searched. The table is effectively compressed and cannot even be executed. Figure 2 is a routing party Schematic diagram of the system logic structure of the application scenario, as shown in Figure 2, the network shown is a very important network in high-performance computers and data centers - fat tree, in which the fat tree network uses destination address routing (Dst -Routing) algorithm, and implementing the routing algorithm by using the above traditional table lookup routing technology to perform route lookup of the data packet. The fat tree Dst-Routing routing algorithm is an Up-Down Routing. The routing rule is: when the source host node sends a data packet to the destination host node, the data packet starts from the source node and continues upward. Routing and forwarding to the nearest public ancestor switching node (switch or router) of the source and destination host switching nodes, then entering the down-route and forwarding phase, routing down and forwarding to the destination host node. When routing up and down, the output port number of the switching node is determined by the destination address of the destination host node and the layer number of the switching node.

Optionally, when the fat tree addresses the switch and the host node according to the method of FIG. 2, if the destination host node network identifier of the data packet is A1.A2.....An, the layer number of the switching node currently arriving by the data packet is L, Then, when the route is up, the port M/2+An-L is selected as the output port number, and when the route is routed, An-L is selected as the output port number, where n is the number of fat tree layers and M is the number of switching node ports.

As shown in Figure 2, when the fat tree Dst-Routing routing algorithm is implemented by using the table lookup routing technology based on the longest prefix matching, it is assumed that the data packet is transmitted from the source node 1.0.0 to the destination node 3.1.1, and the routing path passes through the switching node. 1.1.3, the lookup table of the switching node 1.1.3 is shown in Table 1,

Table 1

网络标识Network identification	掩码Mask	输出端口Output port	…...
网络标识Network identification	掩码Mask	输出端口Output port	…...	1.0.01.0.0	255.255.0255.255.0	00	…...
1.1.01.1.0	255.255.0255.255.0	11	…...	1.0.01.0.0	255.255.0255.255.0	00	…...
1.1.01.1.0	255.255.0255.255.0	11	…...	0.0.00.0.0	255.255.0255.255.0	22	…...
0.1.00.1.0	255.255.0255.255.0	33	…...	0.0.00.0.0	255.255.0255.255.0	22	…...
0.1.00.1.0	255.255.0255.255.0	33	…...	2.0.02.0.0	255.255.0255.255.0	22	…...
2.1.02.1.0	255.255.0255.255.0	33	…...	2.0.02.0.0	255.255.0255.255.0	22	…...
2.1.02.1.0	255.255.0255.255.0	33	…...	3.0.03.0.0	255.255.0255.255.0	22	…...
3.1.03.1.0	255.255.0255.255.0	33	…...	3.0.03.0.0	255.255.0255.255.0	22	…...

Performing a route lookup based on the matching condition "(destination address & mask) == (network identifier & mask)", it can be seen that the logical operation result of the last entry of Table 1 is true, indicating that the matching, the output port of the last entry of the match For port 3, switching node 1.1.3 forwards the packet from output port 3. As can be seen from Table 1, the number of table entries (8) of the lookup table of the network device node <1.1.3> is far more than the number of ports of the switch (4), compared with no compression (16, which is equal to the number of host nodes) ), Only 50% compression is achieved. When the network scale is extended to hundreds of thousands of node sizes, table size and table lookup efficiency will become serious problems. Therefore, using the longest prefix matching table lookup routing technology to implement fat tree Dst-Routing is very inefficient and cannot achieve efficient lookup table compression. The inefficiency of the root cause is the limited ability to express lookup table entries, making table entries not efficiently aggregated.

3 is an exemplary flow diagram of a routing method 300 to achieve more efficient routing forwarding in accordance with an embodiment of the present invention. In a particular implementation, routing method 300 can be performed by, for example, but not limited to, a router or switch. As shown in FIG. 3, the method 300 is specifically as follows:

S302: Receive a data packet, where the data packet carries a destination address of the data packet.

S304: Determine, in the routing table, a matching entry corresponding to the destination address, where the entry of the routing table includes an address operator and a network identifier, where a valid bit of the network identifier of the matching entry is related to the destination address The logical operation result of the valid bit operation according to the address operator of the matching entry is true.

S306: Send the data packet from an output port determined according to output port information in the matching entry.

Optionally, the network identifier includes but is not limited to a MAC address, an IP address, and the like.

Optionally, the address operator includes, but is not limited to, a relational comparison operation equal to (==), not equal to (≠), greater than (>), less than (<), greater than or equal to (≥), and less than or equal to (≤).

Optionally, the valid bit of the network identifier is an overall value that is used to treat the network identifier, and the valid bit of the destination address refers to an overall value that is used as the destination address. For example, the IP address 11.12.14.10 is taken as the value 11121410. Thus, operations can be performed in accordance with the address operator.

Optionally, the entry of the routing table further includes a mask, where a valid bit of the destination address is an overall value of the destination address and the mask bit and the result of the operation, and the network identifier is valid. The bit is an overall value of the network identification and the result of the mask bit and operation.

Optionally, in order to enhance the expressive capability of the network identifier after the mask operation, the mask includes a continuous mask and a non-continuous mask, for example, 255.0.255.0.

Optionally, the comparison terms on both sides of the address operator are each compared as an integer value.

Optionally, the comparison term of the address operator can be from high byte to low word from the first dimension to the nth dimension. The order of the sections is arranged in order from the first dimension to the nth dimension from the low byte to the high byte order.

Optionally, the comparison item of the address operator may further select an m-dimensional label from the n-dimensional label as an overall value, and the m-dimensional label is a high byte to a low word from the first dimension to the mth dimension. The order is arranged in order from the first dimension to the mth dimension in descending order of high byte order.

Optionally, the entry of the routing table further includes priority information, where the priority information is used to indicate a priority of the belonging entry;

Optionally, the priority may be a value to identify the priority of the lookup table entry. The greater the value, the higher the priority, or the smaller the value, the higher the priority. The priority can also be other expressions that can distinguish between size or order, such as English characters.

Optionally, the determining, in the routing table, the matching entry corresponding to the destination address, including: searching for a valid bit of the network identifier of the entry in the routing table according to a priority from highest to lowest, and the destination address. The entry of the valid bit according to the address operator of the entry is a true entry, and the first entry found is used as the match entry, wherein the entries of the lookup table are sorted according to priority.

Optionally, the sending, by the output port determined according to the output port information in the matching entry, the data packet comprises: determining, according to the result of the operation, outputting the port information according to the address information, determining the output port And transmitting the data packet from the output port, the address information comprising at least one of the following: the destination address, a source address of the data packet, a network identifier of the matching entry, and a current node address.

Optionally, the sending the data packet from the output port determined according to the output port information in the matching entry, including: selecting one port from the multiple ports recorded in the output port information according to a hash algorithm Describe the output port and forward the data packet from the output port.

Optionally, the matching entry further includes operation information, where the operation information is used to indicate that the data packet is operated, and the output port determined according to the output port information in the matching entry is forwarded before sending the data packet. Further comprising: operating the data packet according to the operation information in the matching entry.

4 is an exemplary flow diagram of a routing method 400 to enable more efficient routing forwarding in accordance with an embodiment of the present invention. In a particular implementation, routing method 400 can be performed by, for example, but not limited to, a router or switch. As shown in FIG. 4, the method 400 is specifically:

S402: Receive a data packet, where the data packet carries a destination address of the data packet.

S404: Determine a matching entry corresponding to the destination address in the routing table.

S406: Determine, according to the result that the address information is operated according to the output port information in the matching entry, the output port, and send the data packet from the output port, where the address information includes at least one of the following: Describe the destination address, the source address of the data packet, the network identifier of the matching entry, and the current node address.

Optionally, the entry of the routing table includes an address operator and a network identifier, and the logic of the valid bit of the network identifier of the matching entry and the valid bit of the destination address are operated according to the address operator of the matching entry. The result of the operation is true.

Optionally, the valid bit of the network identifier is an overall value that is used to treat the network identifier, and the valid bit of the destination address refers to an overall value that is used as the destination address.

Optionally, in order to enhance the expressive capability of the network identifier after the mask operation, the mask includes Continued mask and non-continuous mask, for example 255.0.255.0.

Optionally, the comparison term of the address operator may be arranged from the first byte to the nth dimension in a high byte to a low byte order or from the first dimension to the nth dimension in a low byte to a high byte order.

Optionally, optionally, the address operator includes, but is not limited to, equal to (==), not equal to (≠), greater than (>), less than (<), greater than or equal to (≥), and less than or equal to (≤), etc. Relationship comparison operation.

According to the technical solution provided by the embodiment, by adding an operation field to the lookup table entry, the expression capability of the lookup table entry is enhanced, and the aggregation capability of the lookup table entry is improved. Thereby reducing The size of the lookup table reduces the on-chip storage requirements, reduces chip area, cost, table lookup time and power consumption, and improves network scale scalability.

FIG. 5 is a schematic structural diagram of a lookup table 500 according to an embodiment of the invention. As shown in FIG. 5, lookup table 500 includes a match field 502 and an action field 504.

Matching field 502 includes, but is not limited to, a network identity and address operator.

Optionally, the network identifier may be represented as an n-dimensional label A1.A2.....An, which may be referred to as a first dimension, a second dimension, ..., an nth dimension from left to right. Each dimension of the network identifier is a numerical value, which can be expressed in any binary form such as binary, decimal, and hexadecimal as needed.

Optionally, the matching field 502 further includes a mask, where the mask is an n-dimensional label corresponding to the network identifier, and the dimension of the dimension and the value of each dimension is the same as the network identifier. The mask performs a binary "bit and (&)" operation with the network identifier to generate an address for performing a lookup table matching. Therefore, it can be considered that a certain bit of a certain dimension value in the mask is 1, and the network identifier corresponds to The binary bit is valid.

Optionally, in order to enhance the expressive capability of the network identifier after the mask operation, the mask field supports both continuous and non-continuous masks, for example, 255.0.255.0.

Optionally, the address operators include, but are not limited to, a relational comparison operation equal to (==), not equal to (≠), greater than (>), less than (<), greater than or equal to (≥), and less than or equal to (≤). The OP is used to perform a comparison operation "destination address OP network identification" or "(destination address & mask) OP (network identification & mask)", where "&" indicates a bit and an operation.

Optionally, the comparison items on both sides of the OP are each compared as a whole value.

Optionally, the comparison items on both sides of the OP may be arranged from the first byte to the nth dimension in a high byte to a low byte order or from the first dimension to the nth dimension in a low byte to a high byte order.

Optionally, the comparison items on both sides of the OP may also respectively select an m-dimensional label from the n-dimensional label as an overall value, and the m-dimensional label is high byte to low byte from the first dimension to the m-th dimension. Arrange sequentially or from low byte to high byte order from the first dimension to the mth dimension.

For example, "210.10.10.0|255.0.255.0|≠" means "the first and third dimension values are not equal to the set of addresses of 210 and 10", so that a certain subset of addresses can be filtered;

"11.12.13.14.15|0.0.0.255.0|==" means that the value of the fourth dimension is 14, which means To compare the value of a dimension in an address;

"11.12.13.14|255.255.255.255|>" indicates an address with a larger address value than 11.12.13.14. Assuming the destination address is "11.12.14.10", the comparison operation performed is "(11.12.14.10&255.255.255.255)>(11.12.13.14&255.255.255.255)", that is, "11.12.14.10>11.12.13.14", The result is "True", which allows you to compare the size of two addresses.

Optionally, the matching field 502 further includes priority information. When the multiple matching entries exist, the matching entry with the highest priority is selected as the final matching entry. The priority can be a value that identifies the priority of the lookup table entry. The greater the value, the higher the priority, or the smaller the value, the higher the priority. The priority can also be other expressions that can distinguish between size or order, such as English characters.

Optionally, all entries of the lookup table 500 are sorted according to priority from high to low. When matching is performed, the matching is performed in descending order of priority.

Optionally, the output port field supports both static and dynamic output port representations. The static port representation is consistent with the traditional lookup table, with a specific constant value representing the port number. The dynamic port representation is based on an algorithm implementation to enhance the expressive power of the output port in the lookup table entry.

Expressions for dynamic output ports include, but are not limited to, the following operations: DimValue(addr,k), FirstDiffDim(addr1, addr2), or FirstDiffDimValue(addr1, addr2).

Where DimValue(addr,k): takes the kth dimension component in the address addr (hereinafter abbreviated as "DV");

FirstDiffDim (addr1, addr2): compares the addresses addr1 and address addr2, and returns the dimension number of the first different dimension component (hereinafter abbreviated as "FDD");

FirstDiffDimValue(addr1, addr2): Compares the addresses addr1 and address addr2, determines the dimension number of the first different dimension, and returns the component of addr1 in this dimension (hereinafter abbreviated as "FDDV").

Optionally, the dynamic port supports address symbol representation, and the address symbol includes but is not limited to: destination address (dst), source address (src), current node address (hst), network identifier (net), and the like. Addr, addr1, and addr2 can each be a destination address (dst), a source address (src), a current node address (hst), or a network identifier (net). Where dst is the destination address in the data packet, src is the source address in the data packet, and hst is the node to which the data packet currently arrives (including The address of the network device, net is the value of the network identification field in the currently processed lookup table entry of the lookup table of the current node.

Optionally, the dynamic port algorithm supports at least one of all mathematical operators such as addition, subtraction, multiplication, division, and shift.

For example, "DV(dst, 2)" means "the second-dimensional address component of the destination address is used as an output port". When dst is 5.6.7.8.9, its output port is 6;

"DV(dst, 4)+4" means "the value of the 4th-order address component in the destination address plus 4 as an output port". When dst is 5.6.7.8.9, the output port is 12.

"FDD(dst, src)" means "the first dimension number different from the dimension component in the destination and source addresses as the output port". When dst is 1.5.7 and src is 1.6.4, the output port is 2.

"FDDV(dst, src)" means "the destination address component of the dimension of the first different dimension component in the destination and source address is taken as the output port". When dst is 1.5.7 and src is 1.6.4, the output port is 5.

Optionally, the output port includes N ports, and any one of the N ports can implement forwarding of the data packet, where N is a natural number greater than 0.

For example, "List(2,4,6)" means "ports 2, 4, and 6"; "Range(6,10)" means "ports 6 to 10".

Optionally, if the output port supports N available ports, the network device selects one forwarding port from the N ports according to a hash algorithm, for forwarding the data packet.

Optionally, the action field 504 also supports other action information for operating on the data packet.

According to the lookup table structure disclosed in this embodiment, by adding an operation field to the lookup table entry, the expression capability of the lookup table entry is enhanced, and the aggregation capability of the lookup table entry is improved. Thereby reducing the size of the lookup table, reducing on-chip storage requirements, reducing chip area, cost, table lookup time and power consumption, and improving network scale scalability.

FIG. 6 is a schematic diagram of a system logical structure of a table lookup routing method application scenario 600 according to an embodiment of the present invention. As shown in FIG. 6, the network shown is a very important network in a high performance computer and a data center - a fat tree. The figure shows the use of the method of the present invention. The fat tree network performs a table lookup routing implementation based on a destination address routing (Dst-Routing) algorithm and a routing forwarding process of a data packet from a source node 1.0.0 to a destination node 3.1.1. The fat tree Dst-Routing routing algorithm is an up/ Up-Down Routing, the routing rule is: When the source host node sends a data packet to the destination host node, the data packet starts from the source node and is always routed up and forwarded to the nearest source and destination host switching node. The common ancestor switches nodes (switches or routers) and then enters the down-route and forwarding phase, routing and forwarding down to the destination host node. When routing up and down, the output port number of the switching node is determined by the network identifier (destination address) of the destination host node and the layer number of the switching node.

Optionally, when the fat tree addresses the switch and the host node according to the manner of FIG. 6, if the destination address of the destination host node of the data packet is A1.A2.....An, the layer number of the switching node currently arriving by the data packet is L. Then, select port M/2+An-L as the output port number when routing upwards, and select An-L as the output port number when routing downwards, where n is the number of fat tree layers and M is the number of switching node ports.

In this embodiment, the end node (host node) address is represented as A1.A2.A3, and the switching node address is represented as B1.B2.(L+M/2), where M is the number of ports of the switching node.

The data packet shall be sent from the source node 1.0.0 to the destination node 3.1.1. When the destination address routing algorithm is adopted, the destination address is routed through the switching nodes 1.0.2, 1.1.3, 1.1.4, 3.1.3 and 3.1. 2. When the routing method according to the present invention performs routing and forwarding of data packets, the lookup table of the switching node 1.0.2 is as shown in Table 2. When the data packet arrives at the switching node 1.0.2, the network device slave data at the switching node 1.0.2 The packet extracts the destination node address, whose destination node address is 3.1.1, and finds the matching entry according to the logical result of "(destination address & mask) OP (network identifier & mask)". If the result is "true", it means matching . The logical operation result of "(3.1.1&255.255.0)==(1.0.0&255.255.0)" is "false", indicating no match; the logic of "(3.1.1&255.255.0)≠(1.0.0&255.255.0)" The result of the operation is "true", indicating a match. The output port of the matching entry is the dynamic output port, DV(dst,3)+2=DV(3.1.1,3)+2=3. According to the dynamic port algorithm and destination address, the final output port is port 3, and the switching node is 1.0. .2 forward the packet from port 3. The next hop switching node connected to port 3 is 1.1.3, so the packet arrives at port 3 of switching node 1.0.2 and arrives at switching node 1.1.3.

Table 2

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	1.0.01.0.0	255.255.0255.255.0	＝＝==	DV(dst，3)DV (dst, 3)	…...
1.0.01.0.0	255.255.0255.255.0	≠≠	DV(dst，3)+2DV(dst,3)+2	…...	1.0.01.0.0	255.255.0255.255.0	＝＝==	DV(dst，3)DV (dst, 3)	…...

The lookup table of the switching node 1.1.3 is shown in Table 3. The logical operation result of "(3.1.1&255.0.0)==(1.0.0&255.0.0)" is "false", indicating no match; "(3.1. The logical operation result of 1&255.0.0)≠(1.0.0&255.0.0)” is “true”, indicating a match. The output port of the matching entry is the dynamic output port, DV(dst, 2)+2=DV(3.1.1, 2) +2=3, according to the dynamic port algorithm and destination address, the final output port is port 3, and the switching node 1.3.1 forwards the packet from port 3. The next hop switching node connected to port 3 is 1.1.4. , so the packet is sent from port 3 of the switching node 1.1.3 to the switching node 1.1.4.

table 3

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	1.0.01.0.0	255.0.0255.0.0	＝＝==	DV(dst，2)DV (dst, 2)	…...
1.0.01.0.0	255.0.0255.0.0	≠≠	DV(dst，2)+2DV(dst,2)+2	…...	1.0.01.0.0	255.0.0255.0.0	＝＝==	DV(dst，2)DV (dst, 2)	…...

The lookup table of the switching node 1.1.4 is shown in Table 4. The logical operation result of "(3.1.1&0.0.0)==(0.0.0&0.0.0)" is "true", indicating a match. The output port of the matching entry is the dynamic output port, DV(dst, 1)=DV(3.1.1,1)=3. According to the dynamic port algorithm and destination address, the final output port is port 3, and the switching node 1.1.4 will be the data. The packet is forwarded from port 3, and the next hop switching node connected to port 3 is 3.1.3, so switching node 1.1.4 forwards the packet from port 3 to switching node 3.1.3.

Table 4

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	0.0.00.0.0	0.0.00.0.0	＝＝==	DV(dst，1)DV(dst,1)	…...

The lookup table of the switching node 3.1.3 is as shown in Table 5. The logical operation result of "(3.1.1&255.0.0)==(3.0.0&255.0.0)" is "true", indicating a match; "((3.1.1&255) The logical operation result of .0.0)≠(3.0.0&255.0.0)” is “false”, indicating no match. The output port of the matching entry is the dynamic output port, DV(dst, 2)=DV(3.1.1, 2) =1, according to the dynamic port algorithm and the destination address, the final output port is port 1, the switching node 3.1.3 forwards the data packet from port 1, and the next hop switching node connected to port 1 is 3.1.2, so the switching node 3.1.3 Forward the packet from port 1 to the switching node 3.1.2.

table 5

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	3.0.03.0.0	255.0.0255.0.0	＝＝==	DV(dst，2)DV (dst, 2)	…...
3.0.03.0.0	255.0.0255.0.0	≠≠	DV(dst，2)+2DV(dst,2)+2	…...	3.0.03.0.0	255.0.0255.0.0	＝＝==	DV(dst，2)DV (dst, 2)	…...

The lookup table of the switching node 3.1.2 is shown in Table 6, "(3.1.1 & 255.255.0) == The logical operation result of (3.1.0&255.255.0)" is "true", indicating a match; "(3.1.1&255.255.0)≠(3.1.0&255.255.0)") has a logical operation result of "false", indicating that it does not match. The output port of the matching entry is a dynamic output port, DV(dst, 3)=DV(3.1.1, 3)=1. According to the dynamic port algorithm and destination address, the final output port is port 1, and the switching node 3.1.2 will The data packet is forwarded from port 1, and the next hop node connected to port 1 is the destination node 3.1.1, so the switching node 1.3.1 forwards the data packet from port 1 to the destination node 3.1.1, thereby completing the packet from the source. Routing and forwarding from node 1.0.0 to destination node 3.1.1.

Table 6

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	3.1.03.1.0	255.255.0255.255.0	＝＝==	DV(dst，3)DV (dst, 3)	…...
3.1.03.1.0	255.255.0255.255.0	≠≠	DV(dst，3)+2DV(dst,3)+2	…...	3.1.03.1.0	255.255.0255.255.0	＝＝==	DV(dst，3)DV (dst, 3)	…...

When the number of layers of the fat tree is n, the destination address (dst) is D1.D2.D3..Dn-L..Dn-1.Dn, the number of ports of the switch node is M, and the layer number of the current switching node is L, currently When the switching node address is C1.C2.....Cn-1.(L+M/2), according to the Dst-Routing routing algorithm, the lookup table of the fat tree top-level switching node can be expressed as the general form shown in Table 7, non- The top-level switching node lookup table can be represented in the general form shown in Table 8. For ease of representation, the mask field is represented here in binary.

Table 7

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	A₁.A₂.….A_n-L-1.0.….0A ₁ .A ₂ .....A _nL-1 .0.....0	11…1.11…1.….11…1.00…0.….00…011...1.11...1.....11...1.00...0.....00...0	＝＝==	DV(dst，n-L)DV (dst, n-L)	…...

Table 8

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	A₁.A₂.….A_n-L-1.0.….0A ₁ .A ₂ .....A _nL-1 .0.....0	11…1.11…1.….11…1.0.0…0.….00…011...1.11...1.....11...1.0.0...0.....00...0	＝＝==	DV(dst，n-L)DV (dst, n-L)	…...
A₁.A₂.….A_n-L-1.0.….0A ₁ .A ₂ .....A _nL-1 .0.....0	11…1.11…1.….11…1.00…0.….00…011...1.11...1.....11...1.00...0.....00...0	≠≠	DV(dst，n-L)+M/2DV(dst,n-L)+M/2	…...	A₁.A₂.….A_n-L-1.0.….0A ₁ .A ₂ .....A _nL-1 .0.....0		＝＝==	DV(dst，n-L)DV (dst, n-L)	…...

According to the routing method disclosed in this embodiment, by adding an operation field to the lookup table entry, the expression capability of the lookup table entry is enhanced, thereby improving the aggregation capability of the lookup table entry. The size of the lookup table is reduced, the requirement for on-chip storage space is reduced, thereby reducing chip area, cost, table lookup time and power consumption, and improving network scale scalability.

FIG. 7 is a schematic diagram of a system logical structure of a routing method application scenario 700 according to an embodiment of the present invention. As shown in FIG. 7, an XY routing algorithm of a two-dimensional mesh network is taken as an example to illustrate an application scenario of the present invention. A mesh network is a direct network of rules. Each node is both a compute node and a switch node. In an n-dimensional mesh network, each node is only associated with it in each dimension. Neighboring (two) nodes are connected. Mesh networks are widely used because of their short connection, easy packaging, good locality, and incremental expansion.

As shown in FIG. 7, taking node 2.2 as an example, according to the present invention, the lookup table is as shown in Table 9.

Table 9

网络标识Network identification	掩码Mask	操作符Operator	优先级priority	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	优先级priority	输出端口Output port	…...	2.02.0	255.0255.0	＜<	1010	00	…...
2.02.0	255.0255.0	＞>	1010	22	…...	2.02.0	255.0255.0	＜<	1010	00	…...
2.02.0	255.0255.0	＞>	1010	22	…...	0.20.2	0.2550.255	＜<	2020	11	…...
0.20.2	0.2550.255	＞>	2020	33	…...	0.20.2	0.2550.255	＜<	2020	11	…...

Based on the routing method disclosed in this embodiment, for a mesh network constructed by any number of nodes and an arbitrary number of dimensions, the number of entries required for the lookup table of any node is not more than the number of ports of the node.

FIG. 8 is a schematic diagram of a system logical structure of a routing method application scenario 800 according to an embodiment of the present invention. As shown in FIG. 8 , a BCube network is taken as an example to illustrate an application scenario of the routing method of the present invention. BCube is a modular solution for large-scale networks. It uses the hyper-cube node connection relationship as a recursive rule. All servers in the same position in different recursive units in the same layer are connected to each other through a switch. This structure requires multiple ports per server, and has the advantages of high hypercube connectivity, small diameter, and high reliability.

When the data packet is transmitted from the source node 1.2 to the destination node 3.3, and sequentially passes through the nodes 2.5, 3.2, and 3.4, and the data packet is forwarded according to the routing method of the present invention, the lookup table of the source node 1.2 is as shown in Table 10, and the purpose of the data packet is The node address is 3.3, and the matching entry is found according to the logical result of "(destination address & mask) OP (network identifier & mask)". If the result is "true", it means matching. Table 10 has only one entry, and the logical operation result of "(destination address &0.0) == (network identifier & 00)" is always "true", indicating a match. The output port of the matching entry is a dynamic output port, 2-FDD(dst, hst)=2-FDD(3.3,1.2)=1. According to the dynamic port algorithm and destination address, the final output port is port 1, and the source node 1.2 will data. The packet is sent from port 1, and the next hop switching node connected to port 1 is 2.5, and the source node 1.2 forwards the packet from port 1 to switching node 2.5.

Table 10

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	0.00.0	0.00.0	＝＝==	2-FDD(dst，hst)2-FDD(dst,hst)	…...

The lookup table of the switching node 2.5 is shown in Table 11, and Table 11 has only one entry, and "( The logical operation result of the address &0.0)==(network identifier &0.0)" is always "true", indicating a match. The output port of the matching entry is a dynamic output port, and DV(dst, 1)=DV(3.3, 1) = 3, according to the dynamic port algorithm and destination address, the final output port is port 3, the switching node 2.5 sends the data packet from port 3, the next hop node connected to port 3 is 3.2, and the switching node 2.5 sends the data packet from Port 3 is forwarded to node 3.2.

Table 11

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	0.00.0	0.00.0	＝＝==	DV(dst，1)DV(dst,1)	…...

The lookup table of node 3.2 is shown in Table 12. Table 12 has only one entry, and the logical operation result of "(destination address &0.0) == (network identifier & 00)" is always "true", indicating a match. The output port of the matching entry is a dynamic output port, 2-FDD(dst, hst)=2-FDD(3.3,3.2)=0. According to the dynamic port algorithm and destination address, the final output port is port 0, and node 3.2 will be the data packet. Forwarded from port 0, the next hop switching node connected to port 0 is 3.4, and node 3.2 forwards the data packet from port 0 to switching node 3.4.

Table 12

The lookup table of the switching node 3.4 is as shown in Table 13. Table 13 has only one entry, and the logical operation result of "(destination address &0.0) == (network identifier & 00)" is always "true", indicating a match. The output port of the matching entry is a dynamic output port, and DV(dst, 2)=DV(3.3, 2)=3. According to the dynamic port algorithm and the destination address, the final output port is port 3, and the switching node 3.4 sends the packet from port 3. issue. The next hop node connected to port 3 is the destination node 3.3, so the switching node 3.4 forwards the data packet from port 3 to the destination node 3.3.

Table 13

网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...
网络标识Network identification	掩码Mask	操作符Operator	输出端口Output port	…...	0.00.0	0.00.0	＝＝==	DV(dst，2)DV (dst, 2)	…...

According to the routing method disclosed in this embodiment, by adding an operation field to the lookup table entry, the expression capability of the lookup table entry is enhanced, thereby improving the aggregation capability of the lookup table entry, and the number of switching nodes and arbitrary network layers for any number of ports. For a BCube network built, only one lookup table entry is required for any node lookup table. Can greatly reduce the need for on-chip storage space, Thereby reducing chip area, cost, table lookup time and power consumption, and improving network scalability.

FIG. 9 is a schematic diagram of a logical structure of a routing device 900 according to an embodiment of the present invention. As shown in FIG. 9, the routing device 900 includes at least a receiving unit 902, a matching unit 904, and a sending unit 906.

The receiving unit 902 is configured to receive a data packet, where the data packet carries a destination address of the data packet;

The matching unit 904 is configured to determine, in the routing table, a matching entry corresponding to the destination address, where the entry of the routing table includes an address operator and a network identifier, where the valid identifier of the network identifier of the matching entry The logical operation result of the valid bit of the destination address being operated according to the address operator of the matching entry is true;

The sending unit 906 is configured to send the data packet from an output port determined according to the output port information in the matching entry.

Optionally, the comparison items on both sides of the address operator may be arranged from the first byte to the nth dimension in a high byte to a low byte order or from the first dimension to the nth dimension in a low byte to a high byte order.

Optionally, the comparison items on both sides of the address operator may each select an m-dimensional label from the n-dimensional label as an overall value, and the m-dimensional label is high byte to low from the first dimension to the mth dimension. The byte order is arranged or from the 1st dimension to the mth dimension in descending order of high byte order.

The matching unit 904 is configured to determine, in the routing table, a matching entry corresponding to the destination address, where the matching unit 904 searches the routing table for the valid bit of the network identifier of the entry and the valid bit of the destination address. The logical operation result of the operation performed by the address operator of the entry is a true entry, and according to the priority information of the found entry, the entry with the highest priority is selected from the searched entry as the match. entry.

Optionally, the matching unit 904 is configured to determine, in the routing table, a matching entry corresponding to the destination address, where the matching unit 904 searches for a network of entries in a routing table according to a priority from highest to lowest priority. An entry of a valid bit of the identifier and a valid bit of the destination address in accordance with an address operator of the entry is a true entry, and the first entry found is the match entry, wherein the lookup The entries for the table are sorted by priority.

Optionally, the sending, by the sending unit 906, the sending, by using the output port determined according to the output port information in the matching entry, the sending, by the sending unit 906, according to the address information, outputting the port information according to the matching entry. As a result of the operation, determining the output port, and transmitting the data packet from the output port, the address information includes at least one of the following: the destination address, a source address of the data packet, and the Match the network ID of the entry and the current node address.

Optionally, the sending, by the sending unit 906, the sending, by the sending unit, the data packet, according to the output port information in the matching entry, includes: the sending unit 906 records from the output port information according to a hash algorithm. One of the plurality of ports is selected as the output port, and the data packet is forwarded from the output port.

Optionally, the device further includes an operation unit, the matching entry further includes operation information, the operation information is used to indicate operation on the data packet, and the operation unit is configured to perform operations according to the matching item. Information, operating on the data packet.

FIG. 10 is a schematic diagram of a logical structure of a routing device 1000 according to an embodiment of the present invention. As shown in FIG. 10, the routing device 1000 includes at least a receiving unit 1002, a matching unit 1004, and a sending unit 1006.

The receiving unit 1002 is configured to receive a data packet, where the data packet carries a destination address of the data packet;

The matching unit 1004 is configured to determine, in the routing table, a matching entry corresponding to the destination address;

The sending unit 1006 is configured to determine, according to the result that the address information is operated according to the output port information in the matching entry, the output port, and send the data packet from the output port, where the address information includes at least the following A: the destination address, a source address of the data packet, a network identifier of the matching entry, and a current node address.

Optionally, the mask includes a continuous mask and a non-continuous mask.

The matching unit 1004 is configured to determine, in the routing table, a matching entry corresponding to the destination address, where the matching unit 1004 searches for a valid bit of the network identifier of the entry in the routing table and a valid bit of the destination address. The logical operation result of the operation performed by the address operator of the entry is a true entry, and according to the priority information of the found entry, the entry with the highest priority is selected from the searched entry as the match. entry.

Optionally, the matching unit 1004 is configured to determine, in the routing table, a matching entry corresponding to the destination address, where the matching unit 1004 searches for a network of entries in a routing table according to a priority from highest to lowest priority. An entry of a valid bit of the identifier and a valid bit of the destination address in accordance with an address operator of the entry is a true entry, and the first entry found is the match entry, wherein the lookup The entries for the table are sorted by priority.

FIG. 11 is a schematic diagram showing the hardware structure of a computing device device 1100 according to an embodiment of the invention. As shown in FIG. 11, computing device 1100 includes a processor 1102, a memory 1104, an input/output interface 1106, a communication interface 1108, and a bus 1110. The processor 1102, the memory 1104, the input/output interface 1106, and the communication interface 1108 implement a communication connection with each other through the bus 1110.

The processor 1102 can be a general-purpose central processing unit (CPU), a microprocessor, and an application-specific integrated circuit (Application Specific Integrated). Circuit, ASIC), or one or more integrated circuits, for performing related procedures to implement the technical solutions provided by the embodiments of the present invention.

The memory 1104 may be a read only memory (ROM), a static storage device, a dynamic storage device, or a random access memory (RAM). The memory 1104 can store an operating system and other applications. When the technical solution provided by the embodiment of the present invention is implemented by software or firmware, the program code for implementing the technical solution provided by the embodiment of the present invention is stored in the memory 1104 and executed by the processor 1102.

The input/output interface 1106 is for receiving input data and information, and outputting data such as operation results.

Communication interface 1108 enables communication between computing device 1100 and other devices or communication networks using transceivers such as, but not limited to, transceivers.

Bus 1110 can include a path for communicating information between various components of computing device 1100, such as processor 1102, memory 1104, input/output interface 1106, and communication interface 1108.

It should be noted that although the computing device 1100 shown in FIG. 11 only shows the processor 1102, the memory 1104, the input/output interface 1106, the communication interface 1108, and the bus 1110, in a specific implementation process, those skilled in the art should It is understood that computing device 1100 also includes other devices necessary to achieve proper operation. Also, those skilled in the art will appreciate that computing device 1100 may also include hardware devices that implement other additional functions, depending on the particular needs. Moreover, those skilled in the art will appreciate that computing device 1100 may also only include the components necessary to implement embodiments of the present invention, and does not necessarily include all of the devices shown in FIG.

The hardware structure shown in FIG. 11 and the foregoing description are applicable to various routing devices provided by the embodiments of the present invention, and are applicable to performing various routing methods provided by the embodiments of the present invention.

In the several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods may be implemented in other manners. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division, and may be implemented in another manner, for example, multiple modules or components may be combined or may be Integrate into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or module, and may be electrical, mechanical or otherwise.

The modules described as separate components may or may not be physically separate. The components displayed as modules may or may not be physical modules, that is, may be located in one place, or may be distributed to multiple network modules. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the embodiment.

In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module. The above integrated modules can be implemented in the form of hardware or in the form of hardware plus software function modules.

The above-described integrated modules implemented in the form of software function modules can be stored in a computer readable storage medium. The software functional modules described above are stored in a storage medium and include instructions for causing a computer device (which may be a personal computer, server, or network device, etc.) to perform some of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: a mobile hard disk, a read-only memory (English: Read-Only Memory, ROM for short), a random access memory (English: Random Access Memory, RAM for short), a magnetic disk or an optical disk, and the like. The medium of the code.

It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and are not limited thereto; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that The technical solutions described in the foregoing embodiments are modified, or some of the technical features are equivalently replaced; and the modifications or substitutions do not deviate from the technical scope of the embodiments of the present invention.

Claims

A routing method, comprising:

Receiving a data packet, where the data packet carries a destination address of the data packet;

Determining, in the routing table, a matching entry corresponding to the destination address, wherein the entry of the routing table includes an address operator and a network identifier, and a valid bit of the network identifier of the matching entry and a valid bit of the destination address The logical operation result of the operation according to the address operator of the matching entry is true;

The data packet is transmitted from an output port determined according to output port information in the matching entry.
The routing method according to claim 1, wherein the valid bit of the network identifier is an overall value that treats the network identifier, and the valid bit of the destination address refers to a destination address. Overall value.
The routing method according to claim 1, wherein the entry of the routing table further includes a mask, and the valid bit of the destination address is the result of the destination address and the mask bit and the operation result. An overall value, the valid bit of the network identifier is an overall value of the network identifier and the result of the mask bit and operation.
The routing method according to claim 3, wherein the mask comprises a continuous mask and a non-continuous mask.
The routing method according to any one of claims 1-4, wherein the entry of the routing table further includes priority information, where the priority information is used to indicate a priority of the belonging entry;

Determining, in the routing table, a matching entry corresponding to the destination address, comprising: a logic for finding a valid bit of the network identifier of the entry in the routing table and a valid bit of the destination address according to an address operator of the entry The operation result is an entry, and according to the priority information of the found entry, the entry with the highest priority is selected from the searched entry as the matching entry.
The routing method according to any one of claims 1 to 4, wherein the determining a matching entry corresponding to the destination address in the routing table comprises: routing the table according to a priority from highest to lowest The entry of the valid bit of the network identifier of the lookup entry and the valid bit of the destination address are logically operated according to the address operator of the entry, and the first entry found is used as the matching entry. Wherein the entries of the lookup table are ranked according to priority sequence.
The routing method according to any one of claims 1 to 6, wherein the transmitting the data packet from an output port determined according to output port information in the matching entry comprises: matching according to the address information And outputting the port information in the entry, determining the output port, and transmitting the data packet from the output port, where the address information includes at least one of the following: the destination address, the data packet Source address, network identifier of the matching entry, and current node address.
The routing method according to any one of claims 1 to 6, wherein the transmitting the data packet from an output port determined according to output port information in the matching entry comprises: according to a hash algorithm One of the plurality of ports recorded in the output port information is selected as the output port, and the data packet is forwarded from the output port.
The routing method according to any one of claims 1 to 8, wherein the matching entry further includes operation information, and the operation information is used to indicate that the data packet is operated according to the matching entry. Before the output port determined by the output port information forwards and sends the data packet, the method further includes: operating the data packet according to the operation information in the matching entry.
The routing method according to any one of claims 1 to 9, wherein the address operator comprises: not equal to, greater than, less than, greater than or equal to, or less than or equal to.
A routing method, comprising:

Receiving a data packet, where the data packet carries a destination address of the data packet;

Determining a matching entry corresponding to the destination address in the routing table;

And determining, according to the result that the address information is operated according to the output port information in the matching entry, the output port, and sending the data packet from the output port, where the address information includes at least one of the following: The address, the source address of the data packet, the network identifier of the matching entry, and the current node address.
The routing method according to claim 11, wherein the entry of the routing table includes an address operator and a network identifier, and a valid bit of the network identifier of the matching entry and a valid bit of the destination address are as described The logical operation of the operation of the matching address's address operator is true.
The routing method according to claim 12, wherein the valid bit of the network identifier is an overall value that treats the network identifier, and the valid bit of the destination address refers to a destination address. Overall value.
The routing method according to claim 12, wherein the entry of the routing table further includes a mask, and the valid bit of the destination address is the result of the destination address and the mask bit and the operation result. An overall value, the valid bit of the network identifier is an overall value of the network identifier and the result of the mask bit and operation.
The routing method according to claim 14, wherein the mask comprises a continuous mask and a non-continuous mask.
The routing method according to any one of claims 12 to 15, wherein the entry of the routing table further includes priority information, where the priority information is used to indicate the priority of the belonging entry;

Determining, in the routing table, a matching entry corresponding to the destination address, comprising: a logic for finding a valid bit of the network identifier of the entry in the routing table and a valid bit of the destination address according to an address operator of the entry The operation result is an entry, and according to the priority information of the found entry, the entry with the highest priority is selected from the searched entry as the matching entry.
The routing method according to any one of claims 12-15, wherein the determining a matching entry corresponding to the destination address in the routing table comprises: routing the table according to a priority from highest to lowest The entry of the valid bit of the network identifier of the lookup entry and the valid bit of the destination address are logically operated according to the address operator of the entry, and the first entry found is used as the matching entry. The entries of the lookup table are sorted according to priority.
The routing method according to any one of claims 11 to 17, wherein the matching entry further includes operation information, and the operation information is used to indicate that the data packet is operated, according to the matching entry. Before the output port determined by the output port information forwards and sends the data packet, the method further includes: operating the data packet according to the operation information in the matching entry.
The routing method according to any one of claims 12 to 18, wherein the address operator comprises: not equal to, greater than, less than, greater than or equal to, or less than or equal to.
A computing device, comprising: a processor, a memory, a bus, and a communication interface;

The memory is for storing a computing device execution instruction, the processor is coupled to the memory via the bus, and when the computing device is running, the processor executes the computer-executed instruction stored by the memory to Causing the computing device to perform the method of any of claims 1-19 method.
A routing device, comprising:

a receiving unit, configured to receive a data packet, where the data packet carries a destination address of the data packet;

a matching unit, configured to determine, in the routing table, a matching entry corresponding to the destination address, where the entry of the routing table includes an address operator and a network identifier, and a valid bit of the network identifier of the matching entry is The logical operation result of the valid bit of the destination address being operated according to the address operator of the matching entry is true;

And a sending unit, configured to send the data packet from an output port determined according to output port information in the matching entry.
The routing device according to claim 21, wherein the valid bit of the network identifier is an overall value that treats the network identifier, and the valid bit of the destination address refers to a destination address. Overall value.
The routing device according to claim 21, wherein the entry of the routing table further includes a mask, and a valid bit of the destination address is a result of the destination address and the mask bit and operation result. An overall value, the valid bit of the network identifier is an overall value of the network identifier and the result of the mask bit and operation.
The routing device of claim 23 wherein the mask comprises a continuous mask and a non-continuous mask.
The routing device according to any one of claims 21 to 24, wherein the entry of the routing table further includes priority information, where the priority information is used to indicate the priority of the belonging entry;

The matching unit is configured to determine, in the routing table, a matching entry corresponding to the destination address, including: the matching unit searches for a valid bit of the network identifier of the entry in the routing table, and a valid bit of the destination address according to the entry The logical operation result of the operation performed by the address operator is a true entry, and based on the priority information of the found entry, an entry having the highest priority is selected from the searched entry as the matching entry.
The routing device according to any one of claims 21 to 24, wherein the matching unit is configured to determine, in the routing table, a matching entry corresponding to the destination address, including: the matching unit according to a priority High to low order in the routing table to find the valid bit of the network identifier of the entry and the valid bit of the destination address in accordance with the address operator of the entry operation logical operation result is true entry, the first found Entries as the matching entry, wherein the The entries of the lookup table are sorted by priority.
The routing device according to any one of claims 21 to 26, wherein the transmitting unit is configured to send the data packet from an output port determined according to output port information in the matching entry, including: the sending The unit determines the output port according to the result of the operation of the output port information in the matching entry according to the address information, and sends the data packet from the output port, where the address information includes at least one of the following: The destination address, the source address of the data packet, the network identifier of the matching entry, and the current node address.
The routing device according to any one of claims 21 to 26, wherein the transmitting unit is configured to send the data packet from an output port determined according to output port information in the matching entry, including: the sending The unit selects one of the plurality of ports recorded in the output port information as the output port according to a hash algorithm, and forwards the data packet from the output port.
The routing device according to any one of claims 21 to 28, wherein the device further comprises an operation unit, the matching entry further comprises operation information, the operation information is used to indicate that the data packet is operated The operation unit is configured to operate the data packet according to the operation information in the matching entry.
The routing device according to any one of claims 21 to 29, wherein the address operator comprises: not equal to, greater than, less than, greater than or equal to, or less than or equal to.
A routing device, comprising:

a receiving unit, configured to receive a data packet, where the data packet carries a destination address of the data packet;

a matching unit, configured to determine, in the routing table, a matching entry corresponding to the destination address;

a sending unit, configured to determine, according to the result that the address information is operated according to the output port information in the matching entry, the output port, and send the data packet from the output port, where the address information includes at least one of the following The destination address, the source address of the data packet, the network identifier of the matching entry, and the current node address.
The routing device according to claim 31, wherein the entry of the routing table includes an address operator and a network identifier, and a valid bit of the network identifier of the matching entry and a valid bit of the destination address are as described The logical operation of the operation of the matching address's address operator is true.
The routing device according to claim 32, wherein the valid bit of the network identifier is an overall value of the network identifier, and the valid bit of the destination address is The destination address is treated as an overall value.
The routing device according to claim 32, wherein the entry of the routing table further includes a mask, and the valid bit of the destination address is the result of the destination address and the mask bit and the operation result. An overall value, the valid bit of the network identifier is an overall value of the network identifier and the result of the mask bit and operation.
The routing device of claim 34 wherein the mask comprises a continuous mask and a non-continuous mask.
The routing device according to any one of claims 32 to 35, wherein the entry of the routing table further includes priority information, where the priority information is used to indicate a priority of the belonging entry;

The matching unit is configured to determine, in the routing table, a matching entry corresponding to the destination address, including: the matching unit searches for a valid bit of the network identifier of the entry in the routing table, and a valid bit of the destination address according to the entry The logical operation result of the operation performed by the address operator is a true entry, and based on the priority information of the found entry, an entry having the highest priority is selected from the searched entry as the matching entry.
The routing device according to any one of claims 32 to 35, wherein the matching unit is configured to determine, in the routing table, a matching entry corresponding to the destination address, including: the matching unit according to a priority High to low order in the routing table to find the valid bit of the network identifier of the entry and the valid bit of the destination address in accordance with the address operator of the entry operation logical operation result is true entry, the first found The entries are the matching entries, wherein the entries of the lookup table are ordered by priority.
The routing device according to any one of claims 31 to 37, wherein the device further comprises an operation unit, the matching entry further comprises operation information, the operation information is used to indicate that the data packet is operated The operation unit is configured to operate the data packet according to the operation information in the matching entry.
The routing device according to any one of claims 32-38, wherein the address operator comprises: equal to, not equal to, greater than, less than, greater than or equal to, or less than or equal to.