CN105099916B - Open flows route exchange device and its processing method to data message - Google Patents

Abstract

The invention discloses an OpenFlow routing switching device and its processing method for data messages. The method includes: responding to receiving an ingress data message, parsing the ingress data message; combining the parsed ingress data message with the The flow entry in the macro flow table is matched; in response to the parsed ingress data packet matching the flow entry in the macro flow table, the start long flow detection associated with the flow entry in the macro flow table is executed action; and in response to the ingress data packet being judged to belong to a long flow in the long flow detection, notify the OpenFlow controller connected to the OpenFlow routing and switching device to detect a long flow, and send a message from the OpenFlow controller The flow entry of the microflow table located in the SRAM is received, and the flow entry of the microflow table is related to the feature value in the ingress data packet corresponding to the detected long flow. The routing switching device and the method thereof reduce the cost of the OpenFlow routing switching device, and enable the OpenFlow controller to optimize the flow forwarding path.

Description

OpenFlow routing and switching equipment and its processing method for data packets

technical field

The present invention relates to network technology, more specifically, to an open flow routing switching device and a processing method for data packets.

Background technique

In the network, Long-lived Flow is also called elephant flow, which refers to the packet flow corresponding to those network connections that survive for a long time and require a large amount of data transmission. For example, FTP downloads a certain video files, the packet flow corresponding to this FTP connection can be regarded as a long flow. Short-lived Flow (Short-lived Flow), also known as mouse flow (mice flow), its characteristics are exactly the opposite of long-lived flow, and refers to the packet flow corresponding to the network connection whose connection survival time is very short and the amount of data is small. For example, a web request usually transmits the data volume of several packets between the client and the server to complete a session, which is a short flow. Recently, with the rapid development of multimedia services and Web2.0, a large number of network measurement and research reports show that 80% of the traffic in the network is contributed by 20% of the long flow. If you can monitor 20% of long flows, you can well control and optimize network traffic, grasp the movement of 80% of the entire network traffic, which will be a very good measure for network traffic optimization.

In traditional communication networks, control logic and data forwarding are tightly coupled to routing and switching devices, resulting in extremely complex management of the network control plane. Software Defined Networking (SDN) technology separates the control and forwarding of the network, and separates the control plane to the core controller or controller cluster, while the routing and switching devices are only responsible for data forwarding. SDN provides a new solution for new network applications and future Internet technologies. Open Flow (OpenFlow) provides technical support for SDN. OpenFlow separates the control function from the network device in the following way: the flow table (Flow Table) structure is maintained on the network device, the message is forwarded according to the flow table entry in the flow table, and the generation, maintenance, and Configuration is managed by the OpenFlow controller. The relationship between SDN and OpenFlow is that SDN is a network design concept, and OpenFlow is a specific technology provided to realize the concept of SDN.

Today, OpenFlow and SDN separate the control and forwarding of network routing and switching devices. This makes the functions of routing and switching devices more single and modular, basically only responsible for packet forwarding. The complex control protocols and logic that were originally integrated into network devices are separated and concentrated on one or more core OpenFlow controllers to manage and control the hardware devices of the entire network. It is very easy to implement network traffic optimization on this centralized network control architecture, and this traffic information is also a good supplement to the OpenFlow controller. Because the OpenFlow controller itself has been able to perceive the topology of the entire network, the most lacking is the real-time dynamic grasp of the entire network traffic information, which requires routing and switching devices to provide information in the network, especially long flow information.

However, the development of OpenFlow technology has encountered a bottleneck that cannot be solved temporarily on the hardware. On the OpenFlow routing and switching device, the above-mentioned flow table is required to process traffic forwarding and define various traffic policies. Existing OpenFlow routing and switching devices generally use TCAM to implement OpenFlow flow tables, because TCAM has the advantages of fast search, simple operation, and support for mask matching, but it has three obvious disadvantages at the same time: high cost, power consumption Yamato table entry update is complex. In addition, the existing OpenFlow routing and switching equipment integrates control devices such as message parsing, matching, and execution, as well as TCAM for implementing flow tables, and even communication devices into the ASIC chip, which is limited by the limitations of ASIC hardware technology, cost, and power consumption. Considering the size and other aspects, ASIC developers generally only integrate a small amount of TCAM in traditional routing and switching equipment, and existing routing and switching equipment cannot monitor long flows. Therefore, there is a need in the art for a low-cost routing and switching device to monitor long flows in the network.

Contents of the invention

According to one aspect of the present invention, a method for processing data packets by an OpenFlow routing and switching device is provided, including:

In response to receiving the ingress data message, parsing the ingress data message;

Matching the parsed ingress data message with the flow entry in the macro flow table of the TCAM, wherein the associated actions in the flow entry of the macro flow table include starting long flow detection;

In response to the parsed ingress data packet matching the flow entry in the macro flow table, perform an action associated with the flow entry in the macro flow table to start long flow detection;

In response to the ingress data packet being judged to belong to a long flow in the long flow detection, the OpenFlow controller connected to the OpenFlow routing and switching device is notified to detect a long flow, and receives a message located in the SRAM from the OpenFlow controller. A flow entry of the microflow table, where the flow entry of the microflow table is related to the feature value in the ingress data packet corresponding to the detected long flow.

According to another aspect of the present invention, an OpenFlow routing and switching device is provided, including:

A parsing device configured to parse the ingress data message in response to receiving the ingress data message;

The matching device is configured to match the parsed ingress data packet with the flow entry in the macro flow table of the TCAM, wherein the associated action in the flow entry of the macro flow table includes starting long flow detection ;

The executing device is configured to execute an action of starting long flow detection associated with the flow entry in the macro flow table in response to the parsed ingress data message matching the flow entry in the macro flow table;

The notification device is configured to, in response to the ingress data message being determined to belong to a long flow in the long flow detection, to notify the OpenFlow controller connected to the OpenFlow routing and switching device The controller receives the flow entry of the microflow table located in the SRAM, and the flow entry of the microflow table is related to the feature value in the entry data packet corresponding to the detected long flow.

The method proposed by the invention well solves the dependence of the ASIC on the TCAM table items, and reduces the requirement of the OpenFlow routing and switching equipment on the TCAM space. By identifying the long flow in the network and migrating the flow entry matching the long flow to the SRAM which is cheaper and has lower energy consumption than TCAM, the cost of the OpenFlow routing and switching equipment is greatly reduced, and an energy-saving design is provided. At the same time, because the present invention identifies the long flow that contributes 80% of the flow, the OpenFlow controller has a further understanding of the flow information of the entire network and finer-grained control, combined with its grasp of the entire network topology In this situation, the traffic forwarding path can be optimized and the occurrence of congestion points can be avoided.

Description of drawings

The above and other objects, features and advantages of the present disclosure will become more apparent by describing the exemplary embodiments of the present disclosure in more detail with reference to the accompanying drawings, wherein, in the exemplary embodiments of the present disclosure, the same reference numerals generally represent same parts.

Figure 1 shows a block diagram of an exemplary computer system/server 12 suitable for use in implementing embodiments of the present invention;

Fig. 2 shows the communication framework of traditional OpenFlow routing switching device and controller;

Fig. 3 shows the flow diagram of the method for processing the data message by the OpenFlow routing and switching equipment; and

FIG. 4 shows the structural composition of an OpenFlow routing and switching device under the OpenFlow communication framework according to an embodiment of the present invention.

Detailed ways

Preferred embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although preferred embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.

Figure 1 shows a block diagram of an exemplary computer system/server 12 suitable for use in implementing embodiments of the present invention. The computer system/server 12 shown in FIG. 1 is only an example, and should not limit the functions and scope of use of the embodiments of the present invention.

As shown in Figure 1, computer system/server 12 takes the form of a general-purpose computing device. Components of computer system/server 12 may include, but are not limited to, one or more processors or processing units 16, system memory 28, bus 18 connecting various system components including system memory 28 and processing unit 16.

Bus 18 represents one or more of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, a processor, or a local bus using any of a variety of bus structures. These architectures include, by way of example, but are not limited to Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, Enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect ( PCI) bus.

Computer system/server 12 typically includes a variety of computer system readable media. These media can be any available media that can be accessed by computer system/server 12 and include both volatile and nonvolatile media, removable and non-removable media.

System memory 28 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) 30 and/or cache memory 32 . Computer system/server 12 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 34 may be used to read and write to non-removable, non-volatile magnetic media (not shown in FIG. 1, commonly referred to as a "hard drive"). Although not shown in Figure 1, a disk drive for reading and writing to removable non-volatile disks (e.g. "floppy disks") may be provided, as well as for removable non-volatile optical disks (e.g. CD-ROM, DVD-ROM) or other optical media) CD-ROM drive. In these cases, each drive may be connected to bus 18 via one or more data media interfaces. Memory 28 may include at least one program product having a set (eg, at least one) of program modules configured to perform the functions of various embodiments of the present invention.

A program/utility 40 having a set (at least one) of program modules 42 may be stored, for example, in memory 28, such program modules 42 including - but not limited to - an operating system, one or more application programs, other program Modules and program data, each or some combination of these examples may include the implementation of the network environment. Program modules 42 generally perform the functions and/or methodologies of the described embodiments of the invention.

Computer system/server 12 may also communicate with one or more external devices 14 (e.g., keyboards, pointing devices, displays 24, etc.), and with one or more devices that enable user interaction with computer system/server 12, And/or communicate with any device (eg, network card, modem, etc.) that enables the computer system/server 12 to communicate with one or more other computing devices. Such communication may occur through input/output (I/O) interface 22 . Also, computer system/server 12 may also communicate with one or more networks (eg, local area network (LAN), wide area network (WAN) and/or public networks such as the Internet) via network adapter 20 . As shown, network adapter 20 communicates with other modules of computer system/server 12 via bus 18 . It should be appreciated that although not shown, other hardware and/or software modules may be used in conjunction with computer system/server 12, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, Tape drives and data backup storage systems, etc.

Figure 2 shows the communication framework of traditional OpenFlow routing and switching devices and controllers. According to Figure 2, under the OpenFlow communication framework, the forwarding and control of traditional routing and switching devices are separated, and OpenFlow routing and switching devices are responsible for datagram Zhang's forwarding, the OpenFlow controller is responsible for the protocol control process. The topology information of the whole network is maintained in the OpenFlow controller. All flow entries in the flow table in the OpenFlow routing and switching device are issued by the OpenFlow controller, and the communication between the OpenFlow routing and switching device and the OpenFlow controller is transmitted on an encrypted channel using the OpenFlow protocol. Each flow entry delivered by the OpenFlow controller can be associated with a statistical counter, and the OpenFlow controller can query the statistical counter of the flow entry to collect traffic statistics.

Several flow tables (Flow Tables) are set on the OpenFlow routing and switching device for storing flow entries related to packet forwarding and policy control. A flow table is usually a collection of a series of flow entries. The OpenFlow specification defines the standard protocol fields that a flow entry needs to support. Each flow entry mainly includes the following information:

1) Matching fields: which fields in the packet header to match, such as source/destination MAC address, VLAN ID, VLAN priority, source/destination IP address, DSCP, IP protocol number, TCP/UDP source/destination port number, etc. . Up to 42 different match fields are defined in the latest OpenFlow specification (openflow-spec-v1.4.0). The flow table has different implementations, and it may only contain certain fields of interest, such as only caring about the IP quintuple; it may also contain most of the fields defined in the OpenFlow specification.

2) Matching priority: each flow entry is assigned a matching priority. A packet may match multiple flow entries in a certain flow table. In this case, only the action associated with the flow entry with the highest matching priority is executed.

3) Associated actions: After the message matches the fields defined in the flow entry, perform the actions associated with it. These actions may include but are not limited to: modify the value of the specified field in the message header, add a new message Text header, delete a certain packet header, send the packet from a certain (certain) port, maintain packet statistics, etc. In addition to these actions, other actions associated with the flow entry can also be customized according to the specific implementation of the flow table.

Referring back to Figure 2, the control devices for packet parsing, matching, and execution in existing OpenFlow routing and switching devices can be decomposed into parsing devices, matching devices, and execution devices. After the ingress data message passes through the parsing device, it will parse out all the fields in the message header, and then the matching device will search in the flow table in which the flow table item is stored in TCAM hardware to see if the message matches the flow table item in it. Matching, if it matches a certain flow entry, the execution device executes the corresponding action in the flow entry. The flow entry in the flow table is issued by the OpenFlow controller to the flow table in the TACM through the communication device of the OpenFlow routing and switching device.

Existing OpenFlow routing and switching devices generally use a ternary content addressable memory (TCAM) as the storage carrier for the flow table. Generally, each bit of the content addressable memory (CAM) has only two values: 0 and 1. Each bit in TCAM has three state values, that is, in addition to 0 and 1, there is also a don't care state, which means that you don't care about the value of this bit, which is often used in the network field. The so-called "mask", for example: an 8-bit binary data 8'b1101xxxx can match all values in the range of 8'b11010000 to 8'b11011111, converted to decimal values, that is, all values between 208 and 223. Because TCAM has high search efficiency and supports the mask mask function of any bit, when TCAM is used as the storage carrier of the flow table, the network management strategy can be implemented very conveniently. For example, if the administrator wants to block all FTP traffic going to 192.168.0.1, it can be expressed under the flow table definition of IP quintuple (source IP address, destination IP address, protocol number, source port number, destination port number) It is (*, 192.168.0, 1, TCP, *, 21), where the symbol * means that the value of this field is not concerned. This kind of flow entry with certain fields masked out usually matches many connections.

The cost of using TCAM is high. In order to monitor long flow at low cost, the present invention further divides the flow table under the traditional OpenFlow framework into macro flow table and micro flow table. Related concepts include: a number of consecutive messages form a message sequence, and one or several message sequences become a data stream. A message sequence between a specific application of two communicating parties is called a microflow. The flow entry in the flow table may require an exact match to all required fields. At this time, the packet sequence matching this exact flow entry is called a "micro flow"; the flow entry of the flow table Certain fields can also be masked out with a "mask". In this case, the sequence of packets matching this mask flow entry is called a "macro flow". Therefore, the flow entry in the macro flow table generally has one or more fields masked out to match the macro flow, and needs to be implemented using TCAM; the flow entry of the micro flow table generally requires an exact match with the preset All fields, i.e. microflows, can be implemented using SRAM. SRAM (Static Random Access Memory) is a static random access memory, which is a kind of memory with a static storage function, which can save the internally stored data without refreshing the circuit. Different from TCAM, each bit in SRAM can only be 0 or 1, so the mask matching function cannot be directly implemented, but the cost of each bit of SRAM is 30 times cheaper than that of TCAM, and the power of each bit Consumption is 150 times lower than TCAM. Dividing the original TCAM-implemented flow table into a TCAM-implemented macro-flow table and an SRAM-implemented micro-flow table can reduce the cost and power consumption of OpenFlow routing and switching devices. The reason for this replacement and how to use it will be introduced later.

On such a premise, the embodiment of the present invention discloses a method for processing data packets by an OpenFlow routing and switching device. FIG. 3 shows a flow chart of a method for processing data packets by an OpenFlow routing and switching device. image 3,

In step S301, in response to receiving the ingress data message, parsing the ingress data message;

In step S302, match the parsed ingress data message with the flow entry located in the macroflow table of the TCAM, wherein the associated action in the flow entry of the macroflow table includes starting long flow detection;

In step S303, in response to the parsed ingress data message matching the flow entry in the macro flow table, execute the action of starting long flow detection associated with the flow entry in the macro flow table;

In step S304, in response to the ingress data packet being judged to belong to a long flow in the long flow detection, the OpenFlow controller connected to the OpenFlow routing and switching device is notified to detect a long flow, and the OpenFlow controller receives The flow entry of the microflow table located in the SRAM is received, and the flow entry of the microflow table is related to the feature value in the ingress data packet corresponding to the detected long flow.

Step S301 is a commonly used technique in existing OpenFlow routing and switching devices, and how to implement it will not be repeated here.

In step S302, the macro-flow table is located in TCAM, and the micro-flow table is located in SRAM. Originally, all flow tables must use TCAM, and the occupied storage space is increased, and the cost and power consumption are high; the micro-flow table implemented by SRAM The flow table replaces the macro flow table implemented by most TCAMs, which can reduce cost and power consumption. A matching field in the flow entry of the macro flow table is masked. For example: flow table definition for IP quintuple (source IP address, destination IP address, protocol number, source port number, destination port number), flow entry (202.197.65.34, 198.54.24.22, TCP, *, 21) It is a flow entry of a macro flow table, indicating all FTP traffic from the host with IP address 202.197.65.34 to the host with IP address 198.54.24.22. Associated actions include actions to enable long-flow detection. When describing the associated actions, it has been pointed out that the associated actions can customize other actions according to the implementation of the flow table. In the present invention, the action of starting long flow detection is defined. This flow entry is also obtained from the OpenFlow controller. In the flow entry of the macro flow table, a flag bit can be defined for a specific field to identify the associated action: whether to start long flow detection, if the flag bit is set (0 or 1), it indicates the associated action The action includes starting long flow detection, if the flag bit is not set (1 or 0), it indicates that the associated action does not include starting long flow detection. For example, add a flag in a certain field: LFD (Long-liVed Flow Detection). If the LFD flag bit is set to 1, it means that when it is detected that an ingress data packet matches the flow entry, the long flow detection mechanism will be started. When the OpenFlow controller sends a flow entry to the macroflow table of the OpenFlow routing and switching device, it can decide whether to set the macroflow flow entry according to the network topology information and the traffic statistics information collected from the routing and switching device. LFD flag. For example, when the controller finds that the traffic destined for a certain network segment is particularly large, it can issue a flow entry to match all the traffic destined for this network segment, and set it in the associated action of the flow entry The LFD flag bit, so as to enable the OpenFlow routing and switching device to recognize the long flow of the packet matching the flow entry.

In step S303, when the ingress data packet matches a certain flow entry in the macro flow table, assuming that the LFD flag is set to enable long flow detection, methods such as probability selection, traffic statistics, and filtering algorithms can be used to judge the flow. Whether the packet belongs to a long flow.

Method 1: Probability selection method

The probabilistic selection method belongs to the prior art. In this method, the flow entry of the macro flow table is associated with the action of probabilistic selection. After the ingress data packet matches the flow entry, it is judged as long flow. Assuming that the selection probability is p, a flow needs to transmit N packets, and each packet has a probability of p to be selected as belonging to the long flow. According to the probability statistics method, the probability that the kth packet of this flow is selected as a long flow is P=1-(1-p)^k. It can be seen from this formula that the larger the value of k, the higher the probability P of the network flow being selected as a long flow. For example, when the probability p=1/64, the probability that a certain flow is judged as a long flow in the 10th packet is 14.6%, and the probability that the 100th packet is judged as a long flow is 79.3%. The probability that the 150th packet is judged as a long flow is as high as 90.1%. This shows that this method has a time process for identifying long streams. Generally speaking, from the perspective of the time span, a long flow needs to transmit a large number of packets, and its probability of being identified is very high, while a short flow is misjudged as a long flow because the number of packets transmitted is very small. probability is very small.

Method 2: traffic statistics method

The traffic statistics method also belongs to the prior art, and the method can identify the message flow corresponding to the connection satisfying a certain flow rate as a long flow. In this method, the flow entry of the macro flow table is associated with a PB (Packet-Byte) counter, and the number of packets (Packet counter) and data volume (Byte counter) are counted at the same time. The PB counter is cleared every fixed period, which can be set by the user. Packets that match the flow entries of the macro flow table will trigger traffic statistics actions: the Packet counter is automatically incremented by 1, and the packet size is added to the Byte counter. Once the Packet counter and Byte counter exceed the specified thresholds at the same time, all packets are selected as long flows until the PB counter is cleared. The screening condition at this stage is to select and determine the packet flow corresponding to the connection meeting a certain rate as a long flow.

Method 3: Filter method

The filter method also belongs to the prior art. In this method, when the packet matches the flow entry of the macro flow table, it enters the Bloom filter for screening. Bloom filters consist of m hash tables, each of which contains N (N is much larger than m) PB counters. The message generates m index values by m hash functions hl, ..., hm, and addresses a PB counter in each hash table. Each PB counter is cleared every other set cycle. Packets that match the flow entries of the macroflow will trigger Bloom filtering actions: all m Packet counters are automatically incremented by 1, and the packet size is added to the m Byte counters respectively. The message is addressed to m PB counters through the Bloom filter. If and only if all the m PB counters exceed the set threshold at the same time, the message is selected as a long flow.

These long-flow identification methods can be classified as probabilistic statistical methods in mathematics, and in principle, a certain amount of samples is required. However, in the network transmission process, the time when packets arrive at the switching device is discrete. Therefore, it takes a certain span of time to collect samples with a large enough capacity. In the process of sample collection, there may be a lagging effect of long-flow identification. For example, a method may recognize a long flow when the 10th packet is transmitted, and the first 9 packets are considered as If it is not a long flow, it will be forwarded through the OpenFlow routing and switching device. The sample size of different methods determines how quickly different long flow identification methods can identify packets as belonging to a certain long flow.

In addition, there are two kinds of misjudgment possibilities in method 1 and method 2: it is possible to misjudge a short flow as a long flow, and it is possible to miss a certain long flow. Method 3 After the parameters are adjusted properly, only short flows are misjudged as long flows. The above three methods can be used alone or combined to obtain higher accuracy and reduce misjudgment.

In step S304, the OpenFlow controller connected to the OpenFlow routing and switching device is notified that the long flow detected can use the existing encrypted channel. Specifically, an interrupt request can be sent to the ASIC hardware driver (communication device) of the OpenFlow routing switching device to tell that a long flow is detected, and the driver sends a PACKET-IN message to the OpenFlow controller. In the PACKET-IN message, Reason The value of the field is set to OFPR_LFI to indicate that the report is determined to belong to a long flow; correspondingly, after the OpenFlow controller learns that the current ingress data packet is a long flow, it extracts a predetermined feature value from the ingress data packet corresponding to the long flow (such as IP quintuples), these feature values are assembled into flow entries of the microflow table, and the OpenFlow controller can, based on a certain policy, Execute actions to modify to achieve its control purpose and strategy, and then send the flow entries of the formed microflow table to the microflow table. Therefore, the flow entry of the microflow table corresponds to the detected long flow The characteristic value in the entry data message is related. Alternatively, the predetermined characteristic value can also be extracted in the OpenFlow routing and switching device, and directly sent to the OpenFlow controller, and then the controller, based on a certain policy, Execute actions to modify, obtain assembled flow table items, and send them to the microflow table.

A direct purpose of the above-mentioned flow entries sent to the micro-flow table is to identify long flows. In the previous steps, long-flow identification has always required the flow entry of the macro flow table; once the characteristic values of the long flow are identified, these characteristic values can be assembled into the flow entry of the micro-flow table, so as to directly identify the flow entries belonging to the long flow table. The packet of the flow can be finely controlled through the associated action, and it is no longer necessary to use the macro flow table to identify the long flow of the data packet belonging to the original long flow. In other words, the work that originally required TACM to be transferred to SRAM can reduce the demand for TCAM. The present invention only needs to identify the long flow on the OpenFlow routing switching device and tell the OpenFlow controller that there may be various control strategies in the OpenFlow controller, and the OpenFlow routing switching device can be set according to these control strategies The flow entry of the microflow table above. Policy enforcement by the OpenFlow controller is outside the scope of this invention.

Therefore, in one embodiment, the method shown in FIG. 3 further includes the following steps (not shown in FIG. 3 ): in step S305, in response to receiving a new entry data message, the new entry data message Match the flow entry in the microflow table.

In another embodiment, the method shown in FIG. 3 further includes (not shown in FIG. 3 ): in step S306, in response to the new entry data packet matching the flow entry in the microflow table, Execute the action associated with the flow entry in the microflow table for the new entry data packet; and in step S307, if the new entry data packet does not match the flow entry in the microflow table , indicating that the entry data packet does not belong to the long flow, then match the new entry data packet with the flow entry in the macro flow table, so as to further determine whether the new entry data packet belongs to the long flow .

Take IP quintuple (source IP address, destination IP address, protocol number, source port number, destination port number) as an example. To monitor the TCP traffic going to the 10.10.1.0/24 network segment, the OpenFlow controller will issue the (*, 10.10.1.0/24, TCP, *, *) table in the macro flow table of the OpenFlow routing and switching device item to monitor these flows and enable long flow detection at the same time. When the message matching the flow entry of the macro flow table is judged to belong to the long flow, the OpenFlow controller extracts the IP quintuple information of the message, assuming (20.20.1.1, 10.10.1.1, TCP, 5001, 21). Furthermore, the OpenFlow controller issues a microflow entry to the microflow table of the OpenFlow routing and switching device, namely (20.20.1.1, 10.10.1.1, TCP, 5001, 21). Then, among the subsequent TCP traffic sent from the host with IP address 20.20.1.1 to the host with IP address 10.10.1.1, all packets with source port number 5001 and destination port number 21 will match the A microflow entry is regarded as a packet belonging to a long flow, and the action to be executed is the action associated with the microflow entry, and the action associated with the matching macroflow entry rule will not be executed. For example, the action to be performed is: transmitting the matching packet from a link with a light load.

In another embodiment, the ingress data packet can also be matched with the flow entries in the microflow table and the macroflow table at the same time. If they are matched at the same time, the matching result with the flow entry in the microflow table is preferred, so that The original long flow and the new long flow can be identified at the same time. That is to say, the method shown in FIG. 3 further includes the following steps (not shown in FIG. 3 ): in step S308, in response to receiving a new entry data packet, simultaneously combine the new entry data packet with the macro flow table Match the flow entry in the micro-flow table; in step S310, in response to the new entry data message at least matching the flow entry in the micro-flow table, execute the micro-flow table entry for the new entry data message The action associated with the flow entry in the macro flow table; in step S311, in response to the new ingress data packet only matching the flow entry in the macro flow table, it is further determined whether the new ingress data packet belongs to a long flow. That is to say, in response to the new ingress data packet matching the flow entries in the macroflow table and the microflow table or only matching the flow entry in the microflow table, the new ingress data packet Execute actions associated with flow entries in the microflow table. This can be achieved by setting the matching priority of the flow entries in the micro-flow table to be higher than that of the flow entries in the macro-flow table, or through an arbitrator.

In one embodiment, when the flow entry of the micro-flow meter is used to judge the discharge, the short flow may be misjudged as the long flow, and these errors need to be eliminated at this time. Therefore, the method shown in FIG. 3 further includes (not shown in FIG. 3 ) when matching the new entry data packet with the flow entry in the microflow table: in step S3011, in response to the new entry The data packet matches the flow entry in the microflow table, and judges whether the new entry data packet is misjudged as belonging to the long flow; Step S3012, in response to determining that the new entry data packet is misjudged as belonging to the long flow flow, notify the OpenFlow controller; step S3013, receive the deleted flow entry of the microflow table from the OpenFlow controller; and step S3014, delete the corresponding flow entry of the microflow table. In this way, the wrongly identified flow entry of the microflow table belonging to the long flow is deleted, and the flow entry of the microflow table is not used to identify the long flow subsequently. The invention also provides a method for judging whether the new entry data message is misjudged as belonging to a long flow. In step 1, associate each flow entry in the microflow table with a counter; in step 2, increase the corresponding counter by 1 in response to the new entry data packet matching the microflow flow entry; Step 3, clear the counter every fixed period; in step 4, when the count before the counter is cleared for several consecutive periods is lower than the set threshold, it is determined that the new entry data message is misjudged as belonging to the long flow,

In one embodiment, after the long-flow data message is sent, the flow entry of the micro-flow table needs to be deleted from the micro-flow table, so that the flow entry of the micro-flow table is not needed to identify the entry data message Whether it belongs to long flow. The above method can be combined, that is to say, the method in FIG. 3 also includes step 312, in response to no data message matching the flow entry of the microflow table in the set period, delete the flow table from the microflow table item. At this point, it means that the long stream has been transmitted and the storage space can be released.

Under the same inventive concept, FIG. 4 shows the structural composition of an OpenFlow routing and switching device 400 under the OpenFlow communication framework according to an embodiment of the present invention. According to FIG. 4 , the OpenFlow routing and switching device includes a micro-flow table 401 , a macro-flow table 402 , an analysis unit 403 , a matching unit 404 , an execution unit 405 and a communication unit 406 . Specifically, in the OpenFlow routing and switching device 400, the parsing means 403 is configured to parse the ingress data message in response to receiving the ingress data message; the matching means 404 is configured to match the parsed ingress data message with the The flow entry in the macroflow table 401 of the TCAM is matched, wherein the associated action in the flow entry of the macroflow table includes starting long flow detection; the executing device 405 is configured to respond to the parsed entry data The message matches the flow entry in the macro flow table, and executes the action of starting long flow detection associated with the flow entry in the macro flow table; the notification device 406 is configured to respond to the entry data message In the long flow detection, it is judged to belong to the long flow, notify the OpenFlow controller connected to the OpenFlow routing and switching device to detect the long flow, and receive the flow table located in the microflow table 402 of the SRAM from the OpenFlow controller item, the flow entry of the microflow table is related to the feature value in the ingress data packet corresponding to the detected long flow.

In one embodiment, the associated action in the flow entry of the macro flow table includes starting long flow detection including: defining a flag bit for a specific field in the flow entry of the macro flow table, wherein the flag bit identifies the associated The action is whether to start long flow detection; in response to the flag bit being set, the associated action in the flow entry of the macro flow table includes

In one embodiment, the execution device 405 executes the action of starting long flow detection associated with the flow entry in the macro flow table by using at least one of the following methods for detection: probabilistic selection; traffic statistics; filtering algorithm .

In one embodiment, the matching device 404 is further configured to: in response to receiving a new ingress data packet, match the new ingress data packet with the flow entry in the microflow table.

In one embodiment, the executing means 405 is further configured to: in response to the new ingress data message matching the flow entry in the microflow table, execute the The action associated with the flow entry in the microflow table; and the matching device 404 is further configured to: respond to the new ingress data packet not matching the flow entry in the microflow table, the The new ingress data packet is matched with the flow entry in the macroflow table, so as to further determine whether the new ingress data packet belongs to a long flow.

In one embodiment, the matching device 404 is further configured to: in response to receiving a new ingress data message, simultaneously combine the new ingress data message with the flow table in the macro flow table and the micro flow table entry; and the executing means 405 is further configured to: in response to the new ingress data message matching at least the flow entry in the micro-flow table, execute the micro-flow table for the new ingress data message The action associated with the flow entry in the macro flow table; in response to the new ingress data packet only matching the flow entry in the macro flow table, further judging whether the new ingress data packet belongs to the long flow.

In one embodiment, the matching device 404 further includes a long flow misjudgment device (not shown in FIG. 4 ), configured to: respond to the new ingress data packet matching the flow entry in the microflow table , judging whether the ingress data packet is misjudged as belonging to the long flow; the notification device is further configured to: in response to determining that the ingress data packet is misjudged as belonging to the long flow, notify the OpenFlow controller; The open flow controller receives the deleted flow entry of the microflow table; and the matching device is further configured to: in response to receiving the deleted flow entry of the microflow table, delete the flow table of the corresponding microflow table item.

In one embodiment, the matching device further includes a long flow misjudgment device configured to: associate each flow entry in the microflow table with a counter; If the flow table item matches, add 1 to the corresponding counter; clear the counter every fixed period; and when the count before the counter is cleared for several consecutive periods is lower than the set threshold, determine that the new entry datagram Wen was misjudged as belonging to Changliu.

In one embodiment, the matching device 404 is further configured to: delete the flow table from the micro-flow table in response to no data packet matching the flow entry of the micro-flow table within the set period item.

The present invention can be a system, method and/or computer program product. A computer program product may include a computer readable storage medium having computer readable program instructions thereon for causing a processor to implement various aspects of the present invention.

A computer readable storage medium may be a tangible device that can retain and store instructions for use by an instruction execution device. A computer readable storage medium may be, for example, but is not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (a non-exhaustive list) of computer-readable storage media include: portable computer diskettes, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or flash memory), static random access memory (SRAM), compact disc read only memory (CD-ROM), digital versatile disc (DVD), memory stick, floppy disk, mechanically encoded device, such as a printer with instructions stored thereon A hole card or a raised structure in a groove, and any suitable combination of the above. As used herein, computer-readable storage media are not to be construed as transient signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (e.g., pulses of light through fiber optic cables), or transmitted electrical signals.

Computer-readable program instructions described herein may be downloaded from a computer-readable storage medium to a respective computing/processing device, or downloaded to an external computer or external storage device over a network, such as the Internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers. A network adapter card or a network interface in each computing/processing device receives computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing/processing device .

Computer program instructions for carrying out operations of the present invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or Source or object code written in any combination, including object-oriented programming languages—such as Java, Smalltalk, C++, etc., and conventional procedural programming languages—such as the “C” language or similar programming languages. Computer readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server implement. In cases involving a remote computer, the remote computer can be connected to the user computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or it can be connected to an external computer (such as via the Internet using an Internet service provider). connect). In some embodiments, an electronic circuit, such as a programmable logic circuit, field programmable gate array (FPGA), or programmable logic array (PLA), can be customized by utilizing state information of computer-readable program instructions, which can Various aspects of the invention are implemented by executing computer readable program instructions.

Aspects of the present invention are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It should be understood that each block of the flowcharts and/or block diagrams, and combinations of blocks in the flowcharts and/or block diagrams, can be implemented by computer-readable program instructions.

These computer-readable program instructions may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus to produce a machine such that when executed by the processor of the computer or other programmable data processing apparatus , producing an apparatus for realizing the functions/actions specified in one or more blocks in the flowchart and/or block diagram. These computer-readable program instructions can also be stored in a computer-readable storage medium, and these instructions cause computers, programmable data processing devices and/or other devices to work in a specific way, so that the computer-readable medium storing instructions includes An article of manufacture comprising instructions for implementing various aspects of the functions/acts specified in one or more blocks in flowcharts and/or block diagrams.

It is also possible to load computer-readable program instructions into a computer, other programmable data processing device, or other equipment, so that a series of operational steps are performed on the computer, other programmable data processing device, or other equipment to produce a computer-implemented process , so that instructions executed on computers, other programmable data processing devices, or other devices implement the functions/actions specified in one or more blocks in the flowcharts and/or block diagrams.

The flowchart and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, a portion of a program segment, or an instruction that includes one or more Executable instructions. In some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks in succession may, in fact, be executed substantially concurrently, or they may sometimes be executed in the reverse order, depending upon the functionality involved. It should also be noted that each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can be implemented by a dedicated hardware-based system that performs the specified function or action , or may be implemented by a combination of dedicated hardware and computer instructions.

Having described various embodiments of the present invention, the foregoing description is exemplary, not exhaustive, and is not limited to the disclosed embodiments. Many modifications and alterations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein is chosen to best explain the principles of the various embodiments, practical applications or technical improvements over technologies in the market, or to enable other persons of ordinary skill in the art to understand the various embodiments disclosed herein.

Claims (18)

1. a kind of open flows route exchange device is to the processing method of data message, including：

In response to receiving entry data message, the entry data message is parsed；

The entry data message of parsing is matched with the flow table item in the macro flow table of TCAM, wherein in the macro stream Associated action flows detection comprising length is started in the flow table item of table；

It is matched with the flow table item in the macro flow table in response to the entry data message of parsing, executes the flow table in the macro flow table The associated action for starting long stream detection of item；

Long stream is belonged to by differentiation in long stream detection in response to the entry data message, notice is set with the open flows route switching Standby connected open flows controller detects long stream, and the flow table of the miniflow table positioned at SRAM is received from the open flows controller , the flow table item of the miniflow table is related to the long characteristic value flowed in corresponding entry data message detected.

2. according to the method described in claim 1, associated action is flowed comprising length is started wherein in the flow table item of the macro flow table Detection includes：Flag bit is defined to specific fields in the flow table item of macro flow table, wherein the associated action of mark bit identification Whether to start long stream detection, it is set in response to the flag bit, then associated action includes in the flow table item of the macro flow table Start long stream detection.

3. according to the method described in claim 1, wherein executing the long stream inspection of the associated startup of flow table item in the macro flow table The action of survey is detected using following at least one method：Probability choosing is sentenced；Traffic statistics；Filter algorithm.

4. according to the method described in claim 1, further including：

In response to receiving new entry data message, by the flow table item in the new entry data message and the miniflow table into Row matching.

5. according to the method described in claim 4, further including：

It is matched with the flow table item in the miniflow table in response to the new entry data message, to this, new entry data message is held The associated action of flow table item in the row miniflow table；

It is mismatched in response to the flow table item in the new entry data message and the miniflow table, by the new entry data message It is matched with the flow table item in the macro flow table, to further differentiate whether the new entry data message belongs to long stream.

6. according to the method described in claim 4, further including：

It matches, judges whether the new entrance number with the flow table item in the miniflow table in response to the new entry data message It is mistaken for belonging to long stream according to message；

In response to determining that the new entry data message is mistaken for belonging to long stream, the open flows controller is notified；

The flow table item for the miniflow table deleted is received from the open flows controller；And

Delete the flow table item of corresponding miniflow table.

7. according to the method described in claim 6, wherein described judge whether to be mistaken for belonging to by the new entry data message Long stream includes：

Every flow table item in miniflow table is associated with a counter；

It is matched with the miniflow flow table item in response to the new entry data message, 1 is added to corresponding counter；

Every the fixed cycle to counter O reset；

When counting before being reset in response to several continuous cycle rate counters is below given threshold, determine the new entrance number It is mistaken for belonging to long stream according to message.

8. according to the method described in claim 1, further including：

In response to receiving new entry data message, while will be in the new entry data message and macro flow table and miniflow table Flow table item is matched；

It is at least matched with the flow table item in miniflow table in response to the new entry data message, to this, new entry data message is held The associated action of flow table item in the row miniflow table；

It is only matched with the flow table item in macro flow table in response to the new entry data message, further differentiates the new entry data Whether message belongs to long stream.

9. according to the method described in claim 1, further including：

In response to the data message not matched with the flow table item of the miniflow table in the setting period, the stream is deleted from miniflow table List item.

10. a kind of open flows route exchange device, including：

Resolver is configured to respond to receive entry data message, parses the entry data message；

Coalignment is configured as the entry data message that will be parsed and the flow table item progress in the macro flow table of TCAM Match, wherein associated action flows detection comprising length is started in the flow table item of the macro flow table；

Executive device, the entry data message for being configured to respond to parsing are matched with the flow table item in the macro flow table, are executed The associated action for starting long stream detection of flow table item in the macro flow table；

It notifies device, is configured to respond to the entry data message and long stream, notice and institute are belonged to by differentiation in long stream detection It states the connected open flows controller of open flows route exchange device and detects long stream, and be located at from open flows controller reception The flow table item of the miniflow table of SRAM, the feature in the flow table item of miniflow table entry data message corresponding with the long stream detected Value is related.

11. open flows route exchange device according to claim 10, wherein associated in the flow table item of the macro flow table Action flows detection comprising startup length：Flag bit is defined to specific fields in the flow table item of macro flow table, wherein the flag bit Associated action is identified whether to start long stream detection；It is set in response to the flag bit, then in the flow table item of the macro flow table Associated action flows detection comprising length is started.

12. open flows route exchange device according to claim 10, wherein the executive device executes the macro flow table In the associated action for starting long stream detection of flow table item be detected using following at least one method：Probability choosing is sentenced；Stream Amount statistics；Filter algorithm.

13. open flows route exchange device according to claim 10, wherein

The coalignment is further configured to：In response to receiving new entry data message, by the new entry data Message is matched with the flow table item in the miniflow table.

14. open flows route exchange device according to claim 13, wherein：

The executive device is further configured to：In response to the flow table item in the new entry data message and the miniflow table Matching, to this, new entry data message executes the associated action of flow table item in the miniflow table；

The coalignment is further configured to：In response to the flow table in the new entry data message and the miniflow table Item mismatches, which is matched with the flow table item in the macro flow table, should to further differentiate Whether new entry data message belongs to long stream.

15. open flows route exchange device according to claim 14, wherein

The coalignment further includes long stream erroneous judgement device, is configured as：In response to the new entry data message with it is described micro- Flow table item matching in flow table, judges whether to be mistaken for the new entry data message to belong to long stream；

The notice device is further configured to：In response to determining that the new entry data message is mistaken for belonging to long Stream, notifies the open flows controller；The flow table item for the miniflow table deleted is received from the open flows controller；

The coalignment is further configured to：Flow table item in response to the miniflow table for receiving deletion is deleted corresponding The flow table item of miniflow table.

16. open flows route exchange device according to claim 15, wherein the coalignment further includes long stream erroneous judgement Device is configured as：

Every flow table item in miniflow table is associated with a counter；

It is matched with the miniflow flow table item in response to the new entry data message, 1 is added to corresponding counter；

Every the fixed cycle to counter O reset；

When counting before being reset in response to several continuous cycle rate counters is below given threshold, determine the new entrance number It is mistaken for belonging to long stream according to message.

17. open flows route exchange device according to claim 10, wherein

The coalignment is further configured to：New enter in response to receiving new entry data message, while by this Mouth data message is matched with the flow table item in macro flow table and miniflow table；

The executive device is further configured to：In response to the new entry data message at least with the flow table item in miniflow table Matching, to this, new entry data message executes the associated action of flow table item in the miniflow table；In response to the entering newly Mouth data message is only matched with the flow table item in macro flow table, further differentiates whether the new entry data message belongs to long stream.

18. open flows route exchange device according to claim 10, wherein the coalignment is also further configured For：In response to the data message not matched with the flow table item of the miniflow table in the setting period, the stream is deleted from miniflow table List item.