WO2016188032A1

WO2016188032A1 - Data forwarding method and system using flow table

Info

Publication number: WO2016188032A1
Application number: PCT/CN2015/093054
Authority: WO
Inventors: 李铭梁
Original assignee: 上海斐讯数据通信技术有限公司
Priority date: 2015-05-26
Filing date: 2015-10-28
Publication date: 2016-12-01
Also published as: CN104954261A; CN104954261B

Abstract

Provided in the invention are a data forwarding method and system using a flow table. The method comprises: presetting, according to a service type of data and by the forwarding system, a virtual flow path comprising at least one flow table; then selecting a virtual flow path according to the service type of received data, selecting a next virtual flow path according to values of each flow table obtained during the flow of the received data in the selected virtual flow path until an attribute of the last flow table in the selected virtual flow path does not allow a goto action; and forwarding the received data according to the values of each flow table in the last virtual flow path. The present invention can effectively prevent an existing OpenFlow switch from performing repeated operations due to the need thereof to output data from a physical output port and then return to a physical input port for several times, thus improving utilization efficiency of a physical port.

Description

Method and system for forwarding data by using flow table

Technical field

The present invention relates to a data forwarding technology, and in particular, to a method and system for forwarding data using a flow table.

Background technique

OpenFlow is a new network exchange model proposed by Stanford University in 2007 to support network innovation research. One of OpenFlow's design goals is to separate the control functions of the network device from the forwarding functions, and then concentrate the control functions on the remote controller. The OpenFlow switch is only responsible for simple and high-speed data forwarding locally. During the operation of an OpenFlow switch, the basis for data forwarding is the flow table. The so-called flow table can be regarded as an abstraction of OpenFlow's data forwarding function for network devices.

In traditional network equipment, TCAM (ternary content addressable memory) is mainly used for FIB, MAC, MPLS Lable and ACL tables. Since the matching fields of each table are of different lengths, they can be designed separately and have a maximum capacity limit in order to achieve the minimum overhead. In the OpenFlow exchange, the FIB, MAC, MPLS Lable table, and the longer ACL table with shorter matching lengths are no longer distinguished, and the flow table entries with the largest length are always used instead. For OpenFlow v1.0, the matching field length of the flow table is up to 252 bits, while the general IPv4RIB TCAM matching field is only 60 to 80 bits long, and its overhead is more than three times.

In order to reduce the flow table overhead, OpenFlow v1.1 designed a multi-level flow table, which is to extract the feature of the flow table, and then break the matching process into multiple steps to form a pipeline processing form, thereby reducing the total number of flow table records. However, the actual physical implementation of multi-level flow meters over three levels is affected by factors such as chip complexity and cost, and the feasibility is relatively low. The maximum number of stream table levels that have been implemented so far is three.

As the complexity of data transmitted in the network increases, when an OpenFlow switch parses data with nested data (such as QinQ data), each layer of parsing must have the next layer of sub-data from the physics of the OpenFlow switch. The output port is then returned to the physical input port of the OpenFlow switch to implement the next layer of nested parsing. Redundant inputs and outputs cause the resources of the physical ports of the OpenFlow switch to become tight. In order to solve the above problems, it is necessary to improve the prior art.

Summary of the invention

In view of the above-mentioned shortcomings of the prior art, the present invention aims to provide a method and system for forwarding data by using a flow table, which is used to solve the problem that the resources of the OpenFlow switch cannot be utilized efficiently in the prior art.

To achieve the above and other related objects, the present invention provides a method for forwarding data using a flow table, including: Presetting a virtual flow path formed by at least one flow table according to the service type of the data; selecting a virtual flow path according to the service type of the received data, and obtaining the flow according to the received data in the selected virtual flow path The value of each flow table is selected, and the next virtual flow path is selected until the attribute of the last flow table in the selected virtual flow path is non-jumpable, according to the value pair of each flow table in the last virtual flow path. The received data is forwarded.

Preferably, the manner of presetting the virtual flow path formed by the at least one flow table according to the service type of the data includes: providing a configuration interface for inputting the service type and the flow table, according to the user in the configuration interface The design operation of the service type and the corresponding virtual flow path generates a virtual flow path.

Preferably, before the step of selecting a virtual flow path according to the service type of the received data, the method further includes: opening a buffer for the received data, and storing the received data into the buffer.

Preferably, the manner of selecting a virtual flow path according to the service type of the received data includes: selecting a virtual flow corresponding to the data in the cache area according to the priority of each preset service type from high to low. path.

Preferably, the virtual flow path includes a port for forwarding data, and correspondingly, when a virtual flow path is selected according to the service type of the received data, and the data is transferred in the selected virtual flow path according to the received data. The value of each flow table obtained is selected, and the next virtual flow path is selected until the attribute of the last flow table in the selected virtual flow path is non-jumpable, according to each flow table in the last virtual flow path. In the process of forwarding the received data, the method further includes: scheduling, according to the preset scheduling policy, different data that needs to use the same flow table and/or port in the respective virtual flow path.

Based on the above object, the present invention further provides a system for forwarding data by using a flow table, comprising: a preset unit, configured to preset a virtual flow path formed by at least one flow table according to a service type of the data; and a forwarding unit, configured to: Selecting a virtual flow path according to the service type of the received data, and selecting a next virtual flow path according to the value of each flow table obtained when the received data flows in the selected virtual flow path until the selected virtual If the attribute of the last flow table in the flow path is non-jumpable, the received data is forwarded according to the value of each flow table in the last virtual flow path.

Preferably, the preset unit is configured to provide a configuration interface for inputting a service type and a flow table, and generating a virtual according to a service type performed by the user in the configuration interface and a corresponding virtual flow path design operation. The flow path.

Preferably, the system further includes: a buffering unit, configured to open a buffer area for the received data and store the received data before the forwarding unit selects a virtual flow path according to the service type of the received data. The buffer area.

Preferably, the forwarding unit is further configured to sequentially select a virtual flow path corresponding to the data in the buffer area according to a preset priority of each service type.

Preferably, the virtual flow path includes a port for forwarding data, and correspondingly, in a process in which the forwarding unit performs forwarding processing for multiple data, the system further includes: a scheduling unit, configured to preset according to the preset The scheduling policy will be scheduled in the respective virtual flow path using different data of the same flow table and/or port.

As described above, the method and system for forwarding data by using the flow table of the present invention have the following beneficial effects: by setting a virtual flow path according to a service type, it is possible to distinguish between data with nested data and non-nested data, and The service type of the data sets the nested virtual flow path of each layer, and determines the virtual flow path corresponding to the next layer of nesting according to the value obtained by matching the data of each flow table in the previous virtual flow path, and directly Go to the corresponding virtual flow path. In this way, it can effectively avoid the repeated operation of the existing OpenFlow switch to output data from the physical output port and then return to the physical input port, which improves the efficiency of the physical port. In addition, the OpenFlow switch can be used. Simultaneously processing multiple virtual flow paths, and each virtual flow path can multiplex the flow tables and ports. Therefore, each flow table and port are scheduled to ensure that each virtual flow path can process the data to be forwarded in time, effectively Improve the efficiency of data forwarding; in addition, provide technicians with a design interface for virtual flow paths, It can facilitate the technician to adjust the existing virtual flow path according to actual needs.

DRAWINGS

1 is a flow chart showing a method of forwarding data using a flow table of the present invention.

FIG. 2 is a schematic diagram showing the flow of data in a plurality of virtual flow paths in the method for forwarding data by using a flow table according to the present invention.

FIG. 3 is a schematic diagram showing the structure of a system for forwarding data by using a flow table according to the present invention.

detailed description

The embodiments of the present invention are described below by way of specific examples, and those skilled in the art can readily understand other advantages and effects of the present invention from the disclosure of the present disclosure. The present invention may be embodied or applied in various other specific embodiments, and various modifications and changes can be made without departing from the spirit and scope of the invention.

As shown in FIG. 1, the present invention provides a method of forwarding data using a flow table. The forwarding method is mainly performed by a forwarding system. The forwarding system includes software and hardware in a network access device including a flow table. The hardware in the network access device includes: a chip for executing the forwarding method. The network access device includes but is not limited to: a switch, a router, and the like.

In step S1, the forwarding system presets a virtual flow consisting of at least one flow table according to the service type of the data. path.

Specifically, the forwarding system saves at least one flow table in advance. The flow table can be regarded as an abstraction of OpenFlow's data forwarding function for network devices. The forwarding system forwards the received data to the at least one flow table according to the virtual flow path, and obtains information such as a destination address, a port, and the like for forwarding the data from the corresponding flow table. The forwarding of one data may be implemented by one of the virtual flow paths, or may be implemented by multiple virtual flow paths. If data is forwarded by the plurality of virtual flow paths, the last flow table in the virtual flow path selected by the forwarding system may jump to the first flow table in the next virtual flow path.

Here, the service type includes, but is not limited to, a video type, a webpage type, a webpage nesting based type, a video nesting based type, and the like. The forwarding system may fixedly set a corresponding virtual flow path according to an existing service type.

Here, each flow table in the virtual flow path has a jumpable/non-jumpable attribute. The flow table whose attribute is non-jumpable is located in the last one of the virtual flow paths.

Preferably, the forwarding system provides a configuration interface for inputting a service type and a flow table, and generates a virtual flow path according to a service type performed by the user in the configuration interface and a corresponding virtual flow path.

Specifically, the forwarding system provides a configuration interface for designing, managing a service type and a virtual flow path. That is, the user can add a flow table to the forwarding system according to the new service type, and design a flow table and a flow table sequence in the corresponding virtual flow path according to the new service class, thereby generating a new virtual flow path.

The forwarding system can also allow the user to change or delete the existing service type and the corresponding virtual flow path to meet the application needs of the switch in different occasions.

In step S2, when the forwarding system receives data, the forwarding system selects a virtual flow path according to the service type of the received data, and obtains the flow according to the received data in the selected virtual flow path. The value of each flow table is selected, and the next virtual flow path is selected until the attribute of the last flow table in the selected virtual flow path is non-jumpable, according to the value of each flow table in the last virtual flow path. The received data is forwarded.

Specifically, the data received by the forwarding system may include one address or multiple addresses. Therefore, when selecting a virtual flow path, the forwarding system needs to select a virtual flow path according to the service type of the data. After the data is sent to the selected virtual flow path, the forwarding system may select whether to further depth the data according to the matching result of each flow table in the virtual flow path and the corresponding protocol in the data. Mining to get the forwarding mechanism and corresponding port optimized for data forwarding.

For example, the data received by the forwarding system belongs to a type based on webpage nesting, and the forwarding system selects a first virtual streaming path corresponding to a type based on webpage nesting. When the forwarding system matches the data according to the webpage transmission protocol provided by each flow table in the selected first virtual flow path, the nested subdata is obtained. The service type is a video type, and the jump instruction of the last flow table in the first virtual flow path is set to the number of the second virtual flow path corresponding to the video type, and the data is sent according to the data. The second virtual flow path. The forwarding system determines, by using the feature matching of each flow table in the second virtual flow path, that the video sub-data in the data has no other nested data in the video sub-data, and then the second The attribute of the last flow table in the virtual flow path is set to be non-jumpable, and the physical output port of the data is determined according to the feature matching result of each flow table in the second virtual flow path and forwarded. The flow process of the first virtual flow path and the second virtual flow path may be referred to FIG. 2 .

Preferably, the forwarding system can perform the flow of the virtual flow path for multiple data at the same time to achieve parallel processing of multiple data. Therefore, before the step of selecting a virtual flow path according to the service type of the received data, the forwarding system further performs: opening a buffer for the received data, and storing the received data into the buffer. step.

For example, the forwarding system can open a buffer pool for the received data and store the received data in the buffer pool. The forwarding system acquires a data from the buffer pool according to the current number of virtual flow paths that are processed in parallel, and allocates a corresponding virtual flow path according to the service type of the acquired data.

More preferably, the forwarding system selects a virtual flow path corresponding to the data in the buffer area according to the priority of each preset service type from high to low.

Specifically, the forwarding system designs the priority of the corresponding virtual flow path according to each service type, and preferentially provides the virtual flow path with the higher priority data in the buffer pool according to the service type of the received data.

For example, the data in the buffer pool is a web page type and a video type, respectively. The virtual flow path of the preset video type of the forwarding system has a higher priority than the virtual flow path of the web page type, and the virtual flow path is preferentially provided according to the data of the video type. The data of the webpage type is still placed in the buffer pool, and after the data processed by the forwarding system in parallel is forwarded, the data of the webpage type is processed.

Since each flow table can be preset in a different virtual flow path, when the forwarding system processes each data in parallel, the same flow table may be used in conflict in different virtual flow paths; or, the forwarding When the system processes each data in parallel, it finally selects the same port for data forwarding. At this time, the forwarding system cannot use one flow table/port for different data at the same time. Therefore, the forwarding system needs to schedule shared resources. Therefore, while the forwarding system performs the step S2, the forwarding system further performs step S3 (not shown).

In step S3, the forwarding system schedules different data of the same flow table and/or port in the respective virtual flow path according to the preset scheduling policy.

Here, the scheduling policy includes but is not limited to at least one of the following: based on strict priority, load balancing, Based on queue scheduling, etc.

For example, the forwarding system monitors that the data type A1 of the webpage type and the data A2 of the video type need to use the flow table B1 in the respective virtual flow path, and according to the strict priority policy, the virtual flow path where the data A2 is located is occupied first. The flow table B1 indicates that the virtual flow path of the data A1 is used to release the flow table B1 when the virtual flow path of the data A2 is released.

As shown in FIG. 3, the present invention provides a system for forwarding data using a flow table. The forwarding system 1 includes software and hardware in a network access device including a flow table. The hardware in the network access device includes: a chip for executing the forwarding method. The network access device includes but is not limited to: a switch, a router, and the like.

The forwarding system 1 includes a preset unit 11 and a forwarding unit 12.

The preset unit 11 is configured to preset a virtual flow path formed by at least one flow table according to a service type of the data.

Specifically, the preset unit 11 pre-stores at least one flow table. The flow table can be regarded as an abstraction of OpenFlow's data forwarding function for network devices. The preset unit 11 streams the received data to the at least one flow table according to the virtual flow path, and can obtain information such as a destination address, a port, and the like for forwarding the data from the corresponding flow table. The forwarding of one data may be implemented by one of the virtual flow paths, or may be implemented by multiple virtual flow paths. If the data is forwarded by the plurality of virtual flow paths, the last flow table in the virtual flow path selected by the preset unit 11 may jump to the first flow table in the next virtual flow path.

Here, the service type includes but is not limited to: a video type, a web page type, and the like. The preset unit 11 can fixedly set a corresponding virtual flow path according to an existing service type.

Preferably, the preset unit 11 provides a configuration interface for inputting a service type and a flow table, and generates a virtual flow path according to a service type performed by the user in the configuration interface and a corresponding virtual flow path design operation.

Specifically, the preset unit 11 provides a configuration interface for designing, managing a service type and a virtual flow path. That is, the user can add a flow table to the preset unit 11 according to the new service type, and design each flow table and flow table sequence in the corresponding virtual flow path according to the new service class, thereby generating a new virtual flow. path.

The preset unit 11 can also allow the user to change or delete the existing service type and the corresponding virtual flow path to meet the application needs of the switch in different occasions.

When the forwarding unit 12 receives data, the forwarding unit 12 is configured to select one virtual flow path according to the service type of the received data, and obtain each of the obtained data in the selected virtual flow path according to the received data. The value of the flow table, select the next virtual flow path until the attribute of the last flow table in the selected virtual flow path is unavailable. If the jump is performed, the received data is forwarded according to the value of each flow table in the last virtual flow path.

Specifically, the data received by the forwarding unit 12 may include one address, and may also include multiple addresses. Therefore, when the virtual flow path is selected, the forwarding unit 12 needs to select a virtual flow path according to the service type of the data. After the data is sent to the selected virtual flow path, the forwarding unit 12 may select whether the data needs to be performed according to the matching result of each flow table in the virtual flow path and the corresponding protocol in the data. Deep mining to get the forwarding mechanism and corresponding port optimized for data forwarding.

For example, the data received by the forwarding unit 12 belongs to a type based on webpage nesting, and the forwarding unit 12 selects a first virtual flow path corresponding to the type of webpage nesting. When the forwarding unit 12 performs feature matching on the data according to the webpage transmission protocol provided by each flow table in the selected first virtual flow path, and obtains the service type of the nested subdata as a video type, Setting a jump instruction of the last flow table in the first virtual flow path to a number of a second virtual flow path corresponding to the video type, and sending the data to the second virtual flow according to the data In the path. The forwarding unit 12 determines, by using the feature matching of each flow table in the second virtual flow path, that the video sub-data in the data has no other nested data in the video sub-data, and then the second The attribute of the last flow table in the virtual flow path is set to be non-jumpable, and the physical output port of the data is determined according to the feature matching result of each flow table in the second virtual flow path and forwarded. The flow process of the first virtual flow path and the second virtual flow path may be referred to FIG. 2 .

Preferably, the forwarding unit 12 can simultaneously perform the flow of the virtual flow path for a plurality of data to achieve the purpose of parallel processing of multiple data. Therefore, before the step of selecting a virtual flow path according to the service type of the received data, the forwarding unit 12 further performs: opening a buffer for the received data, and storing the received data into the buffer. A step of.

For example, the forwarding unit 12 may open a buffer pool for the received data and store the received data in the buffer pool. The forwarding unit 12 acquires a data from the buffer pool according to the current number of virtual flow paths that are processed in parallel, and allocates a corresponding virtual flow path according to the service type of the acquired data.

More preferably, the forwarding unit 12 sequentially selects a virtual flow path corresponding to the data in the buffer area according to the priority of each preset service type.

Specifically, the forwarding unit 12 designs the priority of the corresponding virtual flow path according to each service type, and preferentially provides the virtual flow path with the higher priority data in the buffer pool according to the service type of the received data.

For example, the data in the buffer pool is a web page type and a video type, respectively. The forwarding unit 12 presets the virtual flow path of the video type to be higher than the virtual flow path of the webpage type, and preferentially provides a virtual flow path according to the data of the video type. And the data of the webpage type is still placed in the buffer pool, waiting for the forwarding unit 12 to process in parallel After the data is forwarded, the data of the webpage type is processed.

Since each flow table can be preset in a different virtual flow path, when the forwarding system 1 processes each data in parallel, the same flow table may be used in conflict in different virtual flow paths; or When the forwarding system 1 processes each data in parallel, the same port is finally selected for data forwarding. At this time, the forwarding system 1 cannot use one flow table/port for different data at the same time. Therefore, the forwarding system 1 needs to schedule shared resources. Therefore, while the forwarding unit 12 is executing, the forwarding system 1 further includes a scheduling unit (not shown).

The scheduling unit is configured to schedule, according to a preset scheduling policy, different data that needs to use the same flow table and/or port in the respective virtual flow path.

Here, the scheduling policy includes, but is not limited to, at least one of the following: based on strict priority, load balancing based, queue-based scheduling, and the like.

For example, the scheduling unit monitors that the data type A1 of the webpage type and the data A2 of the video type need to use the flow table B1 in the respective virtual flow path, and according to the strict priority policy, the virtual flow path where the data A2 is located is occupied first. The flow table B1 indicates that the virtual flow path of the data A1 is used to release the flow table B1 when the virtual flow path of the data A2 is released.

In summary, the method and system for forwarding data by using the flow table of the present invention can distinguish between the data with the nested data and the non-nested data by setting the virtual flow path according to the service type, and the service for the nested data. The type sets the nested virtual flow path of each layer, and determines the virtual flow path corresponding to the next layer of nesting according to the value obtained by matching the data of each flow table in the previous virtual flow path, and directly transfers to the corresponding virtual The flow path can effectively avoid the repeated operation of the existing OpenFlow switch to output data from the physical output port and then return to the physical input port, thereby improving the efficiency of using the physical port. In addition, the OpenFlow switch can process multiple pieces at the same time. A virtual flow path, and each virtual flow path can multiplex the flow tables and ports. Therefore, each flow table and port are scheduled to ensure that each virtual flow path can process the data to be forwarded in time, thereby effectively improving data forwarding. Efficiency; in addition, providing the technician with a virtual flow path design interface that facilitates the technician root The existing virtual flow path is adjusted according to the actual demand; and the virtual flow path is adopted, and the design difficulty of the chip can be applied to the four-level or even more flow table without increasing the design difficulty of the chip. Therefore, the present invention effectively overcomes various shortcomings in the prior art and has high industrial utilization value.

The above-described embodiments are merely illustrative of the principles of the invention and its effects, and are not intended to limit the invention. Modifications or variations of the above-described embodiments may be made by those skilled in the art without departing from the spirit and scope of the invention. Therefore, all equivalent modifications or changes made by those skilled in the art without departing from the spirit and scope of the invention will be covered by the appended claims.

Claims

A method for forwarding data by using a flow table, comprising:

Presetting a virtual flow path formed by at least one flow table according to a service type of the data;

Selecting a virtual flow path according to the service type of the received data, and selecting a next virtual flow path according to the value of each flow table obtained when the received data flows in the selected virtual flow path until the selected virtual If the attribute of the last flow table in the flow path is non-jumpable, the received data is forwarded according to the value of each flow table in the last virtual flow path.
The method for forwarding data by using a flow table according to claim 1, wherein the manner of presetting a virtual flow path formed by at least one flow table according to a service type of the data includes:

A configuration interface for inputting a service type and a flow table is provided, and a virtual flow path is generated according to a service type performed by the user in the configuration interface and a corresponding virtual flow path.
The method for forwarding data by using a flow table according to claim 1, wherein before the step of selecting a virtual flow path according to the service type of the received data, the method further comprises: opening a buffer for the received data. And storing the received data into the buffer area.
The method for forwarding data by using a flow table according to claim 3, wherein the manner of selecting a virtual flow path according to the service type of the received data comprises:

According to the preset priority of each service type, the data in the buffer area is sequentially selected to be a virtual flow path.
The method for forwarding data by using a flow table according to claim 1, wherein the virtual flow path includes a port for forwarding data, and correspondingly, selecting a virtual flow path according to a service type of the received data, And selecting, according to the value of each flow table obtained when the received data flows in the selected virtual flow path, selecting a next virtual flow path until the attribute of the last flow table in the selected virtual flow path is non-hopping In the process of forwarding the received data according to the value of each flow table in the last virtual flow path, the method further includes:

According to the preset scheduling policy, different data of the same flow table and/or port need to be scheduled in the respective virtual flow path.
A system for forwarding data by using a flow table, comprising:

a preset unit, configured to preset a virtual flow path formed by the at least one flow table according to the service type of the data;

a forwarding unit, configured to select a virtual flow path according to the service type of the received data, and select a next virtual flow path according to the value of each flow table obtained when the received data flows in the selected virtual flow path, Until the attribute of the last flow table in the selected virtual flow path is non-jumpable, the received data is forwarded according to the value of each flow table in the last virtual flow path.
The system for forwarding data by using a flow table according to claim 6, wherein the preset unit is configured to provide a configuration interface for inputting a service type and a flow table, according to a user in the configuration interface. The design operation of the service type and the corresponding virtual flow path generates a virtual flow path.
The system for forwarding data by using a flow table according to claim 6, further comprising: a buffering unit, configured to receive the virtual flow path before the forwarding unit selects a virtual flow path according to the service type of the received data. The data opens up a buffer area and stores the received data into the buffer area.
The system for forwarding data by using a flow table according to claim 8, wherein the forwarding unit is further configured to sequentially perform data in the buffer area according to a preset priority of each service type from high to low. Corresponding to select a virtual flow path.
The system for forwarding data by using a flow table according to claim 6, wherein the virtual flow path includes a port for forwarding data, and correspondingly, the forwarding unit performs forwarding processing for multiple data. The system further includes: a scheduling unit, configured to schedule, according to a preset scheduling policy, different data that needs to use the same flow table and/or port in the respective virtual flow path.