WO2015027374A1

WO2015027374A1 - Data plane feature configuration method and apparatus

Info

Publication number: WO2015027374A1
Application number: PCT/CN2013/082268
Authority: WO
Inventors: 李彦; 赵广; 唐鹏合
Original assignee: 华为技术有限公司
Priority date: 2013-08-26
Filing date: 2013-08-26
Publication date: 2015-03-05
Also published as: CN104584504B; CN104584504A

Abstract

Embodiments of the present invention relate to a data plane feature configuration method and an apparatus. The method comprises: a data plan receiving a configuration command delivered by a control plane, wherein the configuration command comprises feature data and feature deployment, wherein the feature data comprises an identifier of an atomic rule, an identifier of an atomic action and a processing instruction for performing corresponding matching on the atomic rule and the atomic action; based on the identifier of the atomic rule and the identifier of the atomic action, selecting the atomic rule from rules of the data plane in a centralized manner to form a categorization rule, and selecting the atomic action from actions of the data plane in a centralized manner to form an action set; based on the processing instruction, performing corresponding matching processing on the categorization rule and the action set to form a feature in the data plane; using the feature deployment to configure the feature at an execution time of the data plane so as to perform matching processing on packets received at the execution time, wherein the feature deployment is specifically a configuration relationship between the feature and the execution time.

Description

Description of the characteristics of the data surface configuration method and device

The present invention relates to the field of communications, and in particular, to a method and apparatus for configuring a data plane. Background technique

In the conventional router implementation method, the control plane (English: control plane) sends the data plane (English: da ta plane ) to the data plane, and then the data plane forwards the packet according to the entry.

Because of the strong coupling between the control plane and the data plane, the strong coupling specifically refers to the inseparable and inseparable characteristics of the data plane and the control plane. Therefore, the characteristic configuration of the data plane is fixed, and the user cannot Configured. When a router forwards a packet, its forwarding process is fixed on the control plane. The user cannot extend the forwarding process.

Normally, the data plane and the control plane cannot be completely separated, and the data plane cannot be programmed, resulting in low packet forwarding efficiency. Summary of the invention

The invention provides a method and device for configuring characteristics of a data surface, so as to realize data surface programming, and decompose the forwarding action into multiple characteristics for processing.

In a first aspect, an embodiment of the present invention provides a method for configuring a feature of a data plane, where the method includes:

The data plane receives a configuration command sent by the control plane, where the configuration command includes characteristic data and a feature deployment, where the characteristic data includes an identifier of an atomization rule, an identifier of an atomization action, and atomizing the atomization rule and the atomization rule. The action performs a processing instruction corresponding to the matching; And according to the identifier of the atomization rule included in the characteristic data and the identifier of the atomization action, the data plane selects an atomization rule from a rule set of the data plane to form a classification rule, and the data plane The action set selects the atomization action to form the action set;

And the data plane performs corresponding matching processing with the action set to form a characteristic in the data plane according to the processing indication included in the characteristic data;

Deploying the feature, the data plane configures the feature at an execution timing of the data plane, so that the data plane processes the received packet at the execution timing, and the feature deployment Specifically, the configuration relationship between the characteristics and the execution timing.

In a first possible implementation manner, the configuration command further includes the atomization rule and the atomization action;

According to the characteristic data, the data plane selects an atomization rule from a rule set of the data plane, and forms a classification rule, and selects an atomization action from the action set of the data plane, and further includes: The action is placed in the action set in the data plane.

In a second possible implementation, the execution time is that when the data plane forwards the packet, it is determined whether the packet needs to be matched with the feature, and the judgment is carried according to the Forwarding the message process is strongly related to the key data structure.

In conjunction with the second possible implementation of the first aspect, in a third possible implementation, the data plane configures the feature at an execution timing of the data plane, so that the data plane is in the The processing of the received message at the execution timing is specifically as follows:

Determining, according to a feature of the packet, the data plane to match the feature of the packet with the classification rule configured in the feature at the execution timing;

When the data plane determines that the feature of the packet meets the classification rule, the data plane processes the packet by using the action set corresponding to the classification rule.

In conjunction with the third possible implementation of the first aspect, in a fourth possible implementation, And the matching, by the data plane, the matching of the feature of the packet with the classification rule of the feature configured on the execution timing, according to the feature of the packet, specifically:

And according to the feature of the message, the data plane performs matching processing with the classification rule in the characteristic configured on the execution timing according to the execution timing of the data plane.

With reference to the third possible implementation manner of the first aspect, in a fifth possible implementation, the determining, by the data plane, that the feature of the packet meets the classification rule includes:

And if the feature of the message conforms to each atomization rule in the classification rule, the data plane determines that the feature of the message conforms to the classification rule.

With reference to the third possible implementation manner of the foregoing aspect, in a sixth possible implementation, the processing, by using the action set corresponding to the classification rule, processing the packet includes: using the classification Each of the atomization actions in the set of actions corresponding to the rule processes the message at the execution time.

With reference to the first aspect, or the first, second, third, fourth, fifth, and sixth possible implementation manners of the first aspect, in a seventh possible implementation manner, the classification rule The invention includes a message space feature, a 4-character time feature, a 4-character organization feature, a 4-character content feature, a 4-character traffic feature, and a message environment feature. The action set includes a message feature acquisition class action and a user visible feature. Forwarded decomposition and nested policy support.

In a second aspect, an embodiment of the present invention provides a data plane feature configuration apparatus, where the apparatus includes:

a receiving unit, configured to receive a configuration command sent by the control plane, where the configuration command includes characteristic data and feature deployment, where the characteristic data includes an identifier of an atomization rule, and an identifier of an atomization action

The identification of the atomization action, selecting atomization rules from the rule set of the data surface, forming a classification rule, and selecting an atomization action from the action set of the data surface to form an action set;

a component unit, configured to: according to the processing indication included in the characteristic data, the classification rule Corresponding matching processing with the action set to form a characteristic in the data plane; a configuration unit, configured to use the feature deployment, to configure the feature at an execution timing of the data plane, so as to be in the The received message is processed at the execution timing, and the characteristic deployment is specifically a configuration relationship between the feature and the execution timing.

In a first possible implementation manner, the configuration command received by the receiving unit further includes the atomization rule and the atomization action;

The apparatus further includes: a placing unit for placing the atomization rule into a rule set in the data plane, and placing the atomization action into an action set in the data plane.

With reference to the first possible implementation manner of the second aspect, in a second possible implementation manner, when the configuration unit performs the execution time of configuring the feature on the data plane, specifically, when forwarding the packet And determining whether the packet needs to match the characteristic, and the determining is carried in a key data structure that is strongly related to the forwarding packet process.

With reference to the second possible implementation of the second aspect, in a third possible implementation, the device further includes:

a matching unit, configured to perform, according to a feature of the packet, a matching process between the feature of the packet and the classification rule configured in the feature at the execution timing;

The processing unit is configured to process the packet by using the action set corresponding to the classification rule when determining that the feature of the packet meets the classification rule.

With the third possible implementation of the second aspect, in a fourth possible implementation, the processing unit is specifically configured to: according to the characteristics of the packet, follow the execution timing, and configure The classification rule in the characteristics on the execution timing performs matching processing.

In conjunction with the third possible implementation of the second aspect, in a fifth possible implementation, the matching unit is further configured to: if the feature of the packet meets each atomization rule in the classification rule And determining that the characteristics of the message conform to the classification rule.

With reference to the third possible implementation of the second aspect, in a sixth possible implementation, the processing unit is specifically configured to: use each of the action sets corresponding to the classification rule The atomization action processes the message at the execution timing.

With reference to the second aspect or the first, second, third, fourth, fifth, and sixth possible implementation manners of the first aspect, in the seventh possible implementation manner, the selecting The classification rule consisting of the unit includes the packet space feature, the packet time feature, the packet organization feature, the message content feature, the packet traffic feature, and the message environment feature. Text feature acquisition class action, user visible feature forwarding decomposition and nested policy support.

In a third aspect, an embodiment of the present invention provides a data plane feature configuration apparatus, where the apparatus includes:

Interface

First processor;

Second processor;

a first memory, configured to store program code required by the first processor;

a second memory, configured to store program code required by the second processor;

The interface is configured to receive a packet;

The program code stored by the second memory includes instructions operable to cause the second processor to perform the following process:

And receiving, by the first processor, a configuration command, where the configuration command includes characteristic data and a feature deployment, where the feature data includes an identifier of an atomization rule, an identifier of an atomization action, and the atomization rule and the The atomization action performs a processing instruction corresponding to the matching;

And selecting an atomization rule from the rule set of the data plane according to the identifier of the atomization rule included in the characteristic data and the identifier of the atomization action, forming a classification rule, and selecting an atom from the action set of the data plane Action, forming a set of actions;

And performing, according to the processing indication included in the characteristic data, the classification rule and the action set to perform corresponding matching processing to form a characteristic in the data plane;

Configuring, by using the feature deployment, the feature is configured at an execution timing of the data plane, so as to process the received packet at the execution timing, where the feature deployment is specifically The configuration relationship between the characteristics and the execution timing.

In a first possible implementation manner, the configuration command received by the second processor further includes the atomization rule and the atomization action;

The program code stored by the second memory further includes instructions operable to cause the second processor to perform the following process:

The atomization rules are placed in a rule set in the data plane, and the atomization action is placed in an action set in the data plane.

With reference to the first possible implementation manner of the third aspect, in a second possible implementation manner, the performing, by the second processor, the execution timing of configuring the feature on the data plane refers to when forwarding a packet And determining whether the packet needs to match the characteristic, and the determining is carried in a key data structure that is strongly related to the forwarding packet process.

With reference to the second possible implementation manner of the third aspect, in a third possible implementation, the program code that is stored by the second memory further includes, where the second processor is configured to perform the feature The instructions configured at the execution timing of the data plane to facilitate the matching process on the received message at the execution timing are:

And performing, according to the feature of the packet, a matching process between the feature of the packet and the classification rule configured in the feature at the execution timing;

When it is determined that the feature of the packet meets the classification rule, the packet is processed by using the action set corresponding to the classification rule.

In conjunction with the third possible implementation of the third aspect, in a fourth possible implementation, the program code stored in the second memory is configured to enable the second processor to perform the packet The instructions for performing the matching process with the classification rules in the characteristics configured in the execution timing are:

And according to the characteristics of the message, matching processing is performed with the classification rule among the characteristics configured on the execution timing in accordance with the execution timing of the data plane.

In conjunction with the third possible implementation of the third aspect, in a fifth possible implementation manner, The program code stored in the second memory may be used to cause the second processor to execute an instruction to determine that a feature of the message conforms to the classification rule process is:

And if the feature of the message meets each atomization rule in the classification rule, determining that the feature of the message conforms to the classification rule.

In conjunction with the third possible implementation of the third aspect, in a sixth possible implementation, the program code stored in the second memory is used to enable the second processor to utilize the classification rule The instruction of the action set to process the message is:

The message is processed at the execution timing by using the atomization action in each of the action sets corresponding to the classification rule.

With reference to the first, second, third, fourth, fifth, and sixth possible implementation manners of the third aspect, in a seventh possible implementation, the second memory is stored The classification rule in the program code includes a message space feature, a message time feature, a message organization feature, a content feature, a text traffic feature, and a text environment feature; and the action set includes a text feature acquisition class action, User-visible feature forwarding decomposition and nested policy support.

The method and device for configuring the characteristics of the data plane provided by the embodiment of the present invention use a unified policy to configure the data plane, and use the classification rule and the action set of the characteristics in the data plane to forward and process the packet received on the data plane, thereby realizing the data plane. Programmable, ultimately separating the data plane from the control plane. DRAWINGS

1 is a flowchart of a method for configuring a data plane according to an embodiment of the present invention.

2 is a schematic diagram of a nesting policy according to an embodiment of the present invention;

FIG. 3 is a schematic diagram of an execution timing according to an embodiment of the present invention;

4 is a schematic diagram of modeling a packet forwarding feature according to an embodiment of the present invention;

FIG. 5 is a structural diagram of a device for configuring a characteristic of a data plane according to an embodiment of the present invention;

FIG. 6 is a schematic structural diagram of a hardware configuration of a data plane according to an embodiment of the present invention; FIG. 7 is a schematic diagram of a system for configuring a feature of a data plane according to an embodiment of the present 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 a partial embodiment of the invention, and not all of the embodiments. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.

In order to facilitate the understanding of the present invention, the embodiments of the present invention are not limited to the embodiments of the present invention.

The following is a detailed description of a method for configuring a data plane according to an embodiment of the present invention. FIG. 1 is a flowchart of a method for configuring a data plane according to an embodiment of the present invention. In the embodiment of the present invention, the following steps are performed. The implementation body is a data plane, and the data plane can be configured inside the router by means of a software module. As shown in FIG. 1, the embodiment specifically includes the following steps:

Step 110: The data plane receives a configuration command sent by the control plane, where the configuration command includes the feature data and the feature deployment, where the feature data includes an identifier of the atomization rule, an identifier of the atomization action, and a control plane generation configuration command. The configuration command includes characteristic data and feature deployment. In the embodiment of the present invention, since the control is visible to the user, the control plane can generate a configuration command by receiving an instruction input by the user, and send the configuration command to the data plane.

The data plane receives the configuration command.

Step 120: According to the identifier of the atomization rule included in the characteristic data and the identifier of the atomization action, the data plane selects an atomization rule from a rule set of the data plane, and forms a classification rule, from the The action of the data plane selects the atomization action to form the action set.

Specifically, according to the characteristic data included in the configuration command, the data plane selects atomization rules from the rule set of the data plane to form a classification rule, and selects any atomization action from the action set of the data plane to form an action set. The rule set and the action set are previously configured in the data plane, and the rule Sets, action sets include multiple atomization rules, atomization actions.

In the embodiment of the present invention, the characteristic data includes an atomization rule to be selected, an identification ID of the atomization action, and after the data surface extracts the characteristic data from the configuration command, the atomicization rule to be selected included in the analysis characteristic data, the atom The identification ID of the action, using the atomization rules and the identification ID of the atomization action, selects the corresponding atomization rule and atomization action from the rule set and the action set.

Further, the characteristic data further includes sorting information for sorting the atomization rules in the classification rule, and the data is sorted by the atomization rules constituting the classification rules by using the sorting information.

Step 130: According to the processing indication included in the characteristic data, the data plane performs corresponding matching processing on the classification rule and the action set to form a characteristic in the data plane.

Specifically, according to the characteristic data included in the configuration command, the data plane performs the corresponding processing on the classification rule formed in step 120 and the action set to form a characteristic in the data plane.

In the embodiment of the present invention, the characteristic data further includes a processing indication, where the processing indication specifically refers to an indication that the atomization rule in the classification rule is matched with the atomization action in the action set.

The matching of the atomization rules in the classification rule with the atomization actions in the action set specifically refers to the atomization actions corresponding to each atomization rule, which are sorted according to the order of the atomization rules.

Step 140: Deploying, by using the feature, the data plane configures the feature at an execution timing of the data plane, so that the data plane processes the received packet at the execution timing, where The feature deployment is specifically a configuration relationship between the feature and the execution timing.

Specifically, according to the characteristic deployment included in the configuration command, the data plane configures the characteristics formed in step 130 at the execution timing of the data plane, so that the data plane forwards the received packet at the execution timing.

The feature deployment is specifically a configuration relationship between the feature and the bearer point, and the data plane utilizes the feature deployment to configure the feature formed in step 130 at the execution time of the data plane, that is, the data plane passes Feature deployment, explicitly configuring the composed features at specific execution times.

Optionally, in the step 110 of the embodiment of the present invention, the configuration command issued by the control plane further includes an atomization rule and an atomization action. Before step 120, the data plane includes the atomization rules included in the configuration command, the atomization action into the rule set, and the action set. Through this step, the data plane is formed according to the configuration command issued by the control plane to form a rule set and an action set; and the atomization rules and atomization actions are selected from the rule set and the action set.

Optionally, the execution time in the step 140 of the embodiment of the present invention is specifically: when the data plane forwards the packet, it is determined whether the packet needs to be matched with the configured feature. The basis for the determination is contained in a key data structure that is strongly related to the process of forwarding the message.

In the embodiment of the present invention, the execution timing is also a bearer point, and the bearer point is specifically a key data structure existing in the data plane.

Optionally, the data plane in step 140 of the embodiment of the present invention configures the characteristic at an execution timing of the data plane, so that the data plane processes the received packet at the execution timing. Specifically:

Specifically, the data plane receives the packet to be forwarded, parses the packet, and obtains the feature of the packet, and the data plane matches the feature of the packet with the classification rule configured in the feature at the execution timing; When it is determined that the characteristics of the packet meet the classification rule, the data plane uses the action set corresponding to the classification rule to forward the packet.

Further, according to the feature of the packet, the data plane matching the feature of the packet with the classification rule in the feature configured on the execution timing specifically includes:

According to the characteristics of the message, the data plane is in accordance with the order of execution on the data plane, and the configuration is executed. The classification rules in the characteristics on the line timing are matched.

Determining, by the data plane, that the feature of the packet meets the classification rule includes:

The data plane determines whether the feature of the message conforms to each atomization rule in the classification rule. If the feature of the message conforms to each atomization rule in the classification rule, the data plane determines that the feature of the message conforms to the classification rule. Otherwise, if The characteristics of the message do not conform to each atomization rule in the classification rule, and the data plane determines that the characteristics of the message do not conform to the classification rule.

Further, the process of processing the packet by using the action set corresponding to the classification rule includes:

If the data plane determines that the characteristics of the message conform to the classification rule, the data plane uses each atomization action in the action set corresponding to the classification rule to forward the message at the execution time.

It can be understood that the message is forwarded by the data plane, and the message has a unified structure. In order to implement the data plane, the embodiment of the present invention, therefore, each feature (or forwarding feature) of the message is Modeling, using a unified mechanism to achieve dynamic configuration.

In the embodiment of the present invention, the forwarding process of the packet is decomposed, and each forwarding process of the packet can be decomposed into a combination of different characteristics (that is, a combination of multiple characteristics forms a message for each message. In the forwarding process, the forwarding process of the packet is visible to the user. The feature is a process of packet processing in packet forwarding. And each feature is atomized, that is, the feature cannot be further decomposed.

In the embodiment of the present invention, the feature is divided into three elements: a classification rule (English: c las s if ier ), an action set (English: behavior ), and an execution timing (English: chance ). That is, each attribute consists of a classification rule, a set of actions corresponding to the classification rule, and an execution timing.

The classification rules can be used to determine the rules that the message conforms to and the actions that need to be performed in the action set in the feature. Classification rules can be, but are not limited to, include the following features:

Message space feature (English: space): Interface for receiving or sending packets

Message time characteristics (English: t ime ) : current time period / some time;

Organizational characteristics (English: organiza t ion ): such as a person, an enterprise, etc.; Message content feature (English: content): The data link layer (12) / network layer (L3) defaults to include, the transport layer (L4) as the service attribute of the packet, the packet application type attribute, the message content attribute, The packet security attributes are merged into the packet header or packet data.

Packet traffic characteristics (English: traffic): such as bandwidth attributes, flow policy behavior attributes (for example, delaying the transmission of packets), etc., the bandwidth attributes and behavior attributes are stored in the packet header or message data;

Packet environment characteristics: If the router interface traffic exceeds 10M and the downstream device occupancy rate exceeds 80%, the router communicates with the downstream device to obtain the downstream device occupancy rate.

Further, the classification rules can be further divided into atomization rules. The so-called atomization rules are defined non-separable rules. Different combinations of atomization rules can be combined into different classification rules. Therefore, the classification rules are composed of different atomization rules.

An action set (also referred to as an action action) refers to a specific process of forwarding a message. The action set includes at least a message class acquisition action, a user visible feature forwarding decomposition, and a nested policy support.

• The acquisition class action of the essay feature includes the following features:

Obtaining the characteristics of the packet space: Obtaining the configuration domain, where the configuration domain is the domain where the source device that sends the packet to be forwarded is located;

Get the packet time feature: Get the current time;

Obtain the packet organization characteristics: Obtain the user relationship tree;

Get the content characteristics of the message: three types of "transport layer (L4), message application type, message content,,;

Obtaining packet traffic characteristics: Packets that are measured after the bandwidth is obtained and traffic policing is configured. The packet dyeing behavior refers to the behavior of marking the packet forwarding level when the router forwards the packet. Behavior;

The user-disaggregated feature forwarding decomposition specifically means that each forwarding process of the packet can be decomposed. For the combination of different characteristics, that is, the combination of multiple characteristics, each time the packet is forwarded, the user's forwarding processing of the message is visible.

Nested policy support as an action refers to the implementation of an atomization action corresponding to the atomization rules included in the classification rule in a feature, that is, into another feature, which is usually used for hierarchical service quality (English: Hierarchical Quality of Service, referred to as: HqoS), as shown in Figure 2:

There is a policy relationship between the attributes, which conforms to the atomization rules (messages that pass the Ethernet port (ethO)) included in the classification rules in the feature1, and performs atomization actions corresponding to the atomization rules (acquiring the user tree). Information), and enter the feature feature2. The two atomization rules included in the classification rule in feature2 are: the first atomization rule (person belongs to enterprise 1), the second atomization rule (person belongs to enterprise 2); the atomization action corresponding to the first atomization rule is The feature feature3, the atomization action corresponding to the second atomization rule is the feature feature4, the feature feature3 and the feature feature2 belong to the enterprise 1 nested relationship; the feature feature4 and the feature feature2 belong to the enterprise 2 nested relationship. That is, the feature feature3 and the feature feature4 are nested in the feature feature 2, and the matching process distinguishes packets from the same port that have personal characteristics belonging to different enterprises, and therefore, the characteristics are related, and Non-decomposed. In Figure 2, the message is forwarded by the feature feature1, the feature feature2, the feature feature5, and the feature feature m, but since the feature feature2 is nested with the feature feature 3 and the special '1" element feature4, Ι "Life featurel, special"! Forwarding processing of "feature feature2, special 'f' feature 3 or feature feature4, feature feature5, and feature feature m.

The execution timing is the process of judging whether to match the feature in the process of forwarding processing, that is, whether it matches the classification rule in the feature, and whether the message matches the classification rule in the feature is generally carried in the forwarding process. In the key data structure of the strong correlation, the key data structure exists in any layer of the Open System Interconnection. In the embodiment of the present invention, the execution timing is also a bearer point. As shown in Figure 3, L1i is the physical layer uplink bearer point, L2i is the data link layer uplink bearer point, L3i is the network layer uplink bearer point, L3f network layer forwarding plane bearer point, and L3e is the network layer. The downlink bearer point, L2e is the data link layer downlink direction bearer point, and Lie is the data layer downlink direction load point. User i is the user's uplink load point and usere is the user's downlink load point. L4 is the transport layer direction bearer point, and app is the application direction bearer point. Features are configured on each bearer point. Wherein, Ll i, L2 i, L3 i, L3f, L3e, L2e, Lie, user i and usere are bearer points in the data plane; L4, app are bearer points in the control plane.

After the data plane receives the packet, the data is forwarded to the packet. When the packet passes through the bearer points in the data plane, the data plane determines whether the packet needs to be processed through the feature on the bearer point. If the feature is matched, the feature of the packet is matched with the classification rule of the feature at the bearer point. If the feature is disabled on the bearer point, the packet does not match the feature of the feature on the bearer point. The text is forwarded to the next bearer point.

In the embodiment of the present invention, when the packet passes the L3f bearer point, if the feature is enabled on the L3f, the feature of the packet matches the classification rule of the feature on the bearer point, and the feature of the packet and the bearer point are matched. When the classification rule of the upper attribute matches, the data plane forwards the message by using the action set corresponding to the classification rule. In FIG. 3, the atomization action included in the action set includes forwarding the message to the L4 bearer point in the control plane. , or forward the packet to the L3e bearer.

It should be noted that the user i load point and the usere load point are virtual load points. The user i lifetime may exist in any one of Ll i, L2 i, L3 i; the usere bearer may exist in any one of L3e, L2e, Lie.

FIG. 4 is a schematic diagram of modeling a packet forwarding feature according to an embodiment of the present invention. As shown in Figure 4, the model includes data plane 1 and control plane 2.

The control surface 1 includes a programmable control surface portion 21 and a non-programmable control surface portion 22, and the control surface 2 generates a configuration command according to an instruction input by the user, and issues a configuration command to the data plane 1. The configuration commands include feature data 15 and feature deployment 16.

Data plane 1 includes various characteristics, such as characteristic 10, characteristic 20, and characteristic 30, each characteristic It consists of classification rule 1 1, action set 12, and execution timing.

The classification rule 11 is composed of atomization rules selected from the rule set 13 , and the data plane combines different indivisible atomization rules in a certain order according to the characteristic data 15 to form a classification rule 11 .

The action set 12 is composed of atomization actions selected from the action set 14, and the data plane combines different indivisible atomization actions in a certain order according to the characteristic data 15 to form an action set 12.

The rule set 13 has previously existed in the data plane 1. In an implementation manner, a plurality of atomization rules are included in the rule set 13 , and the data plane is selected according to the characteristic data 15 in the rule set 13 The partial atomization rule constitutes a classification rule in the characteristic, or, in another implementation manner, the data plane puts the atomization rule included in the configuration command into the rule set 13 and selects according to the characteristic data, and constitutes the characteristic Classification rules.

The action set 14 has previously existed in the data plane 1, and in one implementation, the action set 14 includes a plurality of atomization actions, and the data plane selects a partial atomization in the action set 14 according to the feature data 15. The set of actions in the action composition feature, or in another implementation, the data plane puts the atomization action included in the configuration command into the action set 14, and then selects according to the feature data to form the action set in the feature.

The data plane uses the characteristic data 15 to process the classification rule 1 1 and the action set 12 to form the characteristic 10. In the same way, composition characteristics 20, characteristics 30.

The data plane is deployed according to the characteristics. 16 The constituent characteristics are configured on different bearer points of the data plane 1. Therefore, when the packet forwarding process is performed, the processing is performed according to different bearer points. When the feature 10 is configured on the bearer point, the corresponding processing is performed.

Therefore, the user can control the data surface generation characteristics by using the control plane, configure the characteristics on different data points of the data plane through feature deployment, realize the configuration of the data plane using the unified strategy, and the data plane can be programmed, the data plane and the control plane. Separation, using the classification rules and action sets of the characteristics in the data plane to forward the message received on the data plane, and realize the programmable side of the data plane Reason.

In one example, there are three characteristics in the data plane, namely, Feature 1, Feature 2, and Feature 3. The classification rule included in feature 1 is classifierl, classif ierl is composed of the first atomization rule and the second atomization rule, and the action set is actionl; the classification rule of feature 2 is classifie, classifie is composed only of the third atomization rule, action set It is act ion2; the classification rule of feature 3 is classified, the classification is composed of the fourth atomization rule and the fifth atomization rule, and the action set is action3. The data plane is configured at the bearer point 1 configuration 1, the feature 2 is configured at the bearer point 2, and the feature 3 is configured at the bearer point 3.

After the data plane receives the packet, the packet enters the bearer point 1, the bearer point 2, and the bearer point 3. When the packet enters the bearer point 1, the feature 1 of the packet is matched, and the feature of the packet is matched with the classification rule of the feature 1 in the data. The first atomization rule and the second atomization rule included in the classification rule, the data plane forwards the message according to the atomization action included in the action set, that is, performs actionl processing, that is, in FIG. 4, the message The dotted line of the forwarding process is indicated by the characteristic 1.

When the packet is processed by the actionl and enters the bearer 2, the feature 2 is deployed in the bearer 2, and the data is matched with the feature of the feature 2 and the feature of the feature 2 is matched. In the embodiment of the present invention, the feature of the packet is If the classification rule of the feature 2 is not matched, the processing of the action set corresponding to the classification rule is not performed, that is, in FIG. 4, the dotted line of the message forwarding flow is not indicated by the characteristic 2.

When the packet enters the bearer point 3 through the actionl process, the feature 3 is deployed in the bearer point 3, and the feature of the packet is matched with the feature of the feature 3 in the data. In the embodiment of the present invention, the feature matching of the packet is matched. For the classification rule of feature 3, the data plane forwards the message according to the atomization action included in the action set, that is, performs action3 processing, that is, in FIG. 4, the dotted line of the message forwarding flow is indicated by the characteristic 3.

Therefore, according to the feature configuration method of the data plane provided by the embodiment of the present invention, the data plane is configured by using a unified policy, and the data received by the data plane is forwarded by using the classification rule and the action set of the characteristics in the data plane, and the data is realized. Programmable, final data surface and control surface Separation.

The method described in the foregoing embodiments can implement the feature configuration method of the data plane. Correspondingly, the embodiment of the present invention further provides a data plane feature configuration device, which is used to implement the feature configuration method of the data plane in the foregoing embodiment, such as As shown in FIG. 5, the apparatus includes: a receiving unit 510, a selecting unit 520, a component unit 530, and a configuration unit 540.

The receiving unit 51 0 is configured to receive a configuration command issued by the control plane, where the configuration command includes characteristic data and feature deployment, where the feature data includes an identifier of an atomization rule, an identifier of an atomization action, and a a processing indication that the atomization rule is matched with the atomization action; the identifier of the atomization action selects an atomization rule from a rule set of the data surface to form a classification rule 'J, an action from the data surface Centrally select atomic actions to form action sets;

The component unit 530 is configured to perform corresponding matching processing on the classification rule and the action set according to the processing indication included in the characteristic data, to form a characteristic in the data plane;

The configuration unit 540 is configured to use the feature deployment to configure the feature at an execution timing of the data plane, so as to process the received packet at the execution timing, where the feature deployment is specifically The configuration relationship between the characteristics and the execution timing.

The configuration command received by the receiving unit 510 further includes the atomization rule and the atomization action;

The apparatus further includes: a loading unit 550 for placing the atomization rule into a rule set in the data plane, and placing the atomization action into an action set in the data plane.

The configuration unit 540 performs the execution timing of configuring the feature on the data plane. Specifically, when the packet is forwarded, it is determined whether the packet needs to be matched with the feature, and the judgment is carried in the In the key data structure strongly related to the forwarding message process.

The device further includes: a matching unit 560, configured to match, according to the feature of the message, the feature of the message with the classification rule of the feature configured on the execution timing Reason

The processing unit 570 is configured to process the packet by using the action set corresponding to the classification rule when determining that the feature of the packet meets the classification rule.

The processing unit 570 is specifically configured to perform matching processing with the classification rule in the characteristics configured on the execution timing according to the characteristics of the message according to the execution timing.

The matching unit 560 is further configured to: if the feature of the packet meets each atomization rule in the classification rule, determine that the feature of the packet conforms to the classification rule.

The processing unit 570 is specifically configured to process the packet at the execution time by using each of the atomization actions in the action set corresponding to the classification rule.

The classification rule composed by the selecting unit 520 includes a message space feature, a message time feature, a text organization feature, a text content feature, a text traffic feature, and a text environment feature; and the action set consisting of the selection unit includes a message Feature acquisition class actions, user visible feature forwarding decomposition, and nested policy support.

Therefore, according to the data plane characteristic configuration apparatus provided by the embodiment of the present invention, the data plane is configured by using a unified policy, and the data received by the data plane is forwarded by using the classification rule and the action set of the characteristics in the data plane, and the data is realized. The surface is programmable, and finally the separation of the data plane from the control plane is achieved.

In addition, the data plane feature configuration apparatus provided by the embodiment of the present invention may also be implemented as follows to implement the data plane feature configuration method in the foregoing embodiment of the present invention. As shown in FIG. 6, the data plane is The characteristic configuration device includes: an interface 610, a first processor 620, a second processor 630, a first memory 640, and a second memory 650. The system bus 660 is used to connect the interface 610, the first processor 620, the second processor 630, the first memory 640, and the second memory 650.

The interface 610 can be one or more of the following: a network interface controller (English: network interface contro l ler, abbreviated: NIC) providing a wired interface, and an Ethernet NIC, which can provide a copper interface. Fiber optic interface or both copper and fiber optic interfaces; A NIC that provides a wireless interface, such as a wireless local area network (English: wireless local area network, abbreviation: WLAN) NIC.

The first processor 620 is configured to execute a command related to the control plane. Further, the first processor 620 receives an instruction input by the user, generates a configuration command according to the instruction input by the user, and sends the configuration command to the second processor 630;

The second processor 630 is configured to execute a command related to the data plane and perform interactive communication with the first processor 620.

The first memory 640 stores the program code and transmits the stored program code to the first processor 620, which is required by the first processor 620.

The second memory 650 stores the program code, and the storage rule set, the action set, the classification rule, the action set, and the like, and transmits the stored program code to the second processor 630, where the program code is required by the second processor 630. of.

The first/second memory may be a volatile memory (English: volatile memory), such as a random access memory (English: random-access memory, abbreviation: RAM); or a non-volatile memory (English: non-volatile) Memory ) , such as flash memory (English: flash memory), hard disk (English: hard disk drive, abbreviation: HDD) or solid state drive (English: solid-state drive, abbreviation: SSD); or a combination of the above types of memory.

The interface 610 in the feature configuration device of the data plane is configured to receive a message.

The program code stored by the second memory 640 includes instructions that can be used to cause the second processor 630 to perform the following process:

And selecting an atomization rule from the rule set of the data plane according to the identifier of the atomization rule included in the characteristic data and the identifier of the atomization action, forming a classification rule, and selecting an atom from the action set of the data plane Action, forming a set of actions; And performing, according to the processing indication included in the characteristic data, matching the classification rule and the action set to form a characteristic in the data plane;

Configuring, by the feature deployment, the feature is configured at an execution timing of the data plane, so as to process the received packet at the execution timing, where the feature deployment is specifically The configuration relationship between the execution timings.

Further, the configuration command received by the second processor 630 further includes the atomization rule and the atomization action;

The program code stored by the second memory 650 also includes instructions that can be used to cause the second processor 630 to perform the following process:

Further, the performing, by the second processor 630, the execution timing of configuring the feature on the data plane, when determining whether the packet needs to be matched with the feature, when the packet is forwarded, the determining According to the key data structure that is strongly related to the process of forwarding the message.

Further, the program code stored by the second memory 650 further includes an operation timing that is operable to cause the second processor 630 to perform configuring the feature at the data plane, so as to be at the execution timing The instruction to perform the matching process on the received message is:

Further, the program code stored by the second memory 650 can be used to cause the second processor 630 to perform the classification of the feature of the message and the characteristic configured on the execution timing. The instructions for the rule to perform the matching process are:

And performing, according to the feature of the message, a matching process with the classification rule in the characteristic configured on the execution timing according to the execution timing of the data plane. Further, the program code stored by the second memory 650 can be used to enable the second processor 630 to execute an instruction that determines that the feature of the message conforms to the classification rule process:

Further, the program code stored by the second memory 650 can be used to enable the second processor 630 to process the message by using the action set corresponding to the classification rule:

Further, the classification rule in the program code stored by the second memory 650 includes a packet space feature, a packet time feature, a packet organization feature, a message content feature, a packet traffic feature, and a message environment. The action set includes a message feature acquisition class action, a user visible feature forwarding decomposition, and a nested policy support.

In addition, the embodiment of the present invention further provides a data surface feature configuration system. As shown in FIG. 7, the system includes the message processing device provided by the foregoing embodiment.

Further, the system further includes: a control plane, wherein the control plane includes a control unit, and the control unit is implemented by a processor.

In FIG. 7, the control unit is configured to send a configuration command to the characteristic configuration device of the data plane, and the receiving unit in the characteristic configuration device of the data plane receives the configuration command delivered by the control unit, where the configuration command is Including atomization rules, atomization actions, feature data, and feature deployment, and transferring configuration commands to the selection unit, component unit, configuration unit, and into the unit, placing the unit into the unit according to the configuration command, placing the atomization rules included in the configuration command Into the rule set, put the atomization action In the action set; the selection unit selects the atomization rule from the rule set according to the configuration command, and uses the characteristic data included in the configuration command to form a classification rule, and selects an atomization action from the action set to form an action set; and the component is classified according to the characteristic data. The rule and the action set are processed correspondingly to form a characteristic; the configuration unit is deployed according to the characteristics included in the configuration command, and the feature is configured at the execution timing of the data plane.

Therefore, according to the data plane feature configuration system provided by the embodiment of the present invention, the data plane is configured by using a unified policy, and the data received by the data plane is forwarded by using the classification rule and the action set of the characteristics in the data plane, and the data is realized. The surface is programmable, and finally the separation of the data plane from the control plane is achieved.

Those skilled in the art should further appreciate that the elements and steps of the various examples described in connection with the embodiments disclosed herein can be implemented by computer software, which has been generally described in terms of function in the above description. Example composition and steps. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the solution. Those skilled in the art can use different methods to implement the described functions for each particular application, but such implementation should not be considered to be beyond the scope of the present invention.

The steps of a method or algorithm described in connection with the embodiments disclosed herein can be implemented by a software module executed by a processor. The software module can be placed in a random access memory (English: RAM), read only memory (English: ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, or the like. Any other form of storage medium.

The above described embodiments of the present invention are further described in detail, and the embodiments of the present invention are intended to be illustrative only. The scope of the protection, any modifications, equivalent substitutions, improvements, etc., which are made on the basis of the technical solutions of the present invention, are included in the scope of the present invention.

Claims

claims

1. A data plane characteristic configuration method, characterized in that the method includes:

The data plane receives the configuration command issued by the control plane. The configuration command includes characteristic data and characteristic deployment. The characteristic data includes the identification of the atomization rule, the identification of the atomization action and the combination of the atomization rule and the atomization The action performs corresponding matching processing instructions;

According to the identification of the atomization rule and the identification of the atomization action included in the characteristic data, the data plane selects the atomization rule from the rule set of the data plane to form a classification rule. The action set selects atomized actions to form an action set;

According to the processing instructions included in the characteristic data, the data plane performs corresponding matching processing on the classification rules and the action set to form characteristics in the data plane;

Utilizing the characteristic deployment, the data plane configures the characteristic at the execution timing of the data plane, so that the data plane processes the received message at the execution timing, and the characteristic deployment Specifically, it is the configuration relationship between the characteristics and the execution timing.

2. The characteristic configuration method of the data plane according to claim 1, characterized in that, the configuration is based on the characteristic data, and the data plane selects atomization rules from the rule set of the data plane to form a classification The rule selects an atomic action from the action set of the data plane, and before forming the action set, it also includes: placing the sub-action into the action set in the data plane.

3. The data plane characteristic configuration method according to claim 1, wherein the execution timing refers to when the data plane forwards a message to determine whether the message needs to match the characteristic, The basis for the judgment is carried in a key data structure that is strongly related to the message forwarding process.

4. The characteristic configuration method of the data plane according to claim 3, characterized in that the data plane configures the characteristics at the execution timing of the data plane, so that the data plane can The details of processing the received message at the execution timing are as follows:

According to the characteristics of the message, the data plane matches the characteristics of the message with the classification rules configured in the characteristics of the execution timing;

When the data plane determines that the characteristics of the message conform to the classification rule, the data plane uses the action set corresponding to the classification rule to process the message.

5. The data plane characteristic configuration method according to claim 4, characterized in that, according to the characteristics of the message, the data plane combines the characteristics of the message with the characteristics configured on the execution timing. The matching processing of the classification rules in the characteristics specifically includes:

According to the characteristics of the message, the data plane performs matching processing with the classification rules in the characteristics configured on the execution timing according to the order of the execution opportunities on the data plane.

6. The data plane characteristic configuration method according to claim 4, wherein the data plane determines that the characteristics of the message comply with the classification rules specifically includes:

If the characteristics of the packet conform to each atomization rule in the classification rules, the data plane determines that the characteristics of the packet conform to the classification rules.

7. The data plane characteristic configuration method according to claim 4, wherein the processing of the message using the action set corresponding to the classification rule specifically includes:

Using each of the atomic actions in the action set corresponding to the classification rule, the message is processed at the execution timing.

8. The data plane characteristic configuration method according to any one of claims 1 to 7, characterized in that the classification rules include message space characteristics, message time characteristics, message organization characteristics, and message content characteristics. Message traffic characteristics and message environment characteristics; The action set includes message characteristic acquisition actions, user-visible characteristic forwarding decomposition and nested policy support.

9. A data plane characteristic configuration device, characterized in that the device includes:

A receiving unit, configured to receive configuration commands issued by the control plane. The configuration commands include feature data and feature deployment. The feature data includes the identifier of the atomization rule and the identifier of the atomization action. identify the atomic action, select the atomic rule from the rule set of the data plane to form a classification rule, select the atomic action from the action set of the data surface to form an action set;

A component unit configured to perform corresponding matching processing on the classification rule and the action set according to the processing instructions included in the characteristic data, to form characteristics in the data plane;

A configuration unit configured to utilize the feature deployment and configure the feature at the execution timing of the data plane so as to process the received message at the execution timing. The feature deployment is specifically: The configuration relationship between the above characteristics and the execution timing.

10. The data plane characteristic configuration device according to claim 9, wherein the configuration command received by the receiving unit further includes the atomization rule and the atomization action; the device further includes: A putting unit is used to put the atomization rule into a rule set in the data plane, and put the atomization action into an action set in the data plane.

1 1. The device for configuring data plane characteristics according to claim 9, wherein the execution timing of the configuration unit configuring the characteristics on the data plane specifically refers to, when forwarding a message, determining Whether the message needs to match the characteristics, the basis for the judgment is carried in a key data structure that is strongly related to the message forwarding process.

12. The data plane characteristic configuration device according to claim 11, characterized in that the device further includes:

A matching unit, configured to match the characteristics of the message with the classification rules configured in the characteristics of the execution timing according to the characteristics of the message;

A processing unit, configured to use the action set corresponding to the classification rule to process the message when it is determined that the characteristics of the message comply with the classification rule.

13. The device for configuring characteristics of the data plane according to claim 12, characterized in that the processing unit is specifically configured to, according to the characteristics of the message, according to the order of the execution timing, and configure the Matching processing is performed on the classification rules in the characteristics on the execution time.

14. The data plane characteristic configuration device according to claim 12, characterized in that: The matching unit is also configured to determine that the characteristics of the message comply with the classification rules if the characteristics of the message comply with each atomization rule in the classification rules.

15. The data plane characteristic configuration device according to claim 12, wherein the processing unit is specifically configured to use each of the atomized actions in the action set corresponding to the classification rule, in The execution timing processes the message.

16. The data plane characteristic configuration device according to any one of claims 9 to 15, characterized in that the classification rules composed of the selection unit include message space characteristics, message time characteristics, and message organization characteristics. , message content characteristics, message flow characteristics and message environment characteristics; the action set composed of the selection unit includes message feature acquisition actions, user-visible feature forwarding decomposition and nested policy support.

17. A data plane characteristic configuration device, characterized in that the device includes: an interface;

first processor;

second processor;

A first memory used to store program codes required by the first processor;

A second memory used to store program codes required by the second processor;

The interface is used to receive messages;

The program code stored in the second memory includes instructions that may be used to cause the second processor to perform the following processes:

Receive a configuration command issued by the first processor. The configuration command includes characteristic data and characteristic deployment. The characteristic data includes an identifier of an atomization rule, an identifier of an atomization action, and a combination of the atomization rule and the atomization rule. Atomized actions perform corresponding matching processing instructions;

According to the identification of the atomization rule and the identification of the atomization action included in the characteristic data, select an atomization rule from the rule set of the data plane to form a classification rule, and select atoms from the action set of the data plane. Transform actions into action sets;

According to the processing instructions included in the characteristic data, combine the classification rule and the action set The corresponding matching processing is performed together to form the characteristics in the data surface;

The characteristic deployment is used to configure the characteristic at the execution timing of the data plane so as to process the received message at the execution timing. The characteristic deployment is specifically the characteristics and The configuration relationship between the execution timings.

18. The data plane characteristic configuration device according to claim 17, wherein the configuration command received by the second processor further includes the atomization rule and the atomization action;

The program code stored in the second memory also includes instructions that may be used to cause the second processor to perform the following processes:

Put the atomization rules into the rule set in the data plane, and put the atomization actions into the action set in the data plane.

19. The device for configuring characteristics of the data plane according to claim 17, characterized in that, the execution timing of the second processor to configure the characteristics on the data plane refers to when forwarding a message, determining the Whether the message needs to match the characteristics, the basis for the judgment is carried in a key data structure that is strongly related to the process of forwarding the message.

20. The device for configuring characteristics of the data plane according to claim 19, wherein the program code stored in the second memory further includes a program code that can be used to cause the second processor to execute configuring the characteristics at the location. At the execution timing of the data plane, the instruction to perform matching processing on the received message at the execution timing is:

According to the characteristics of the message, matching the characteristics of the message with the classification rule in the characteristics configured on the execution timing;

When it is determined that the characteristics of the message conform to the classification rule, the action set corresponding to the classification rule is used to process the message.

21. The device for configuring characteristics of the data plane according to claim 20, wherein the program code stored in the second memory can be used to cause the second processor to execute the configuration of the characteristics of the message. The classification rules in the characteristics on the execution timing are matched. The process instructions are:

According to the characteristics of the message, in accordance with the order of the execution timing on the data plane, matching processing is performed with the classification rules in the characteristics configured on the execution timing.

22. The device for configuring characteristics of the data plane according to claim 20, characterized in that the program code stored in the second memory can be used to cause the second processor to execute the determination that the characteristics of the message meet the requirements. The instructions for the classification rule process are:

If the characteristics of the message comply with each atomization rule in the classification rules, it is determined that the characteristics of the message comply with the classification rules.

23. The data plane characteristic configuration device according to claim 20, wherein the program code stored in the second memory can be used to cause the second processor to use the action corresponding to the classification rule. The instructions for gathering and processing the packets are:

24. The data plane characteristic configuration device according to any one of claims 17 to 23, wherein the classification rules in the program code stored in the second memory include message space characteristics, Text time characteristics, text organization characteristics, text content characteristics, text flow characteristics and text environment characteristics; the action set includes message feature acquisition actions, user-visible feature forwarding decomposition and nested strategies support.