Network optimization method and system for high-frequency transaction

Abstract

The invention relates to the field of electronic commerce, in particular to a high-frequency transaction-oriented network optimization method and system, and the method specifically comprises the steps: setting a high-frequency transaction traffic recognition unit in an OVS instance, and enabling the high-frequency transaction traffic recognition unit to recognize high-frequency transaction traffic and forward a data packet to a DPDK data plane of the OVS instance, otherwise, forwarding to the kernel data plane of the OVS instance; the DPDK data plane judges whether the data packet is matched with a flow table adopted by the DPDK data plane or not, if not, the data packet is forwarded to the kernel data plane, if the kernel data plane does not have the matched flow table, a Packet-In event is triggered, and a control plane of the OVS instance processes the data packet and generates a new flow table item; the flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency transaction flow identification unit and the flow table of the kernel data plane, and the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance.

Description

Technical Field

[0001] The present invention relates to the field of e-commerce, and in particular to a network optimization method and system for high-frequency transactions. Background Art

[0002] With the continuous development of e-commerce and financial markets, the speed of high-frequency trading is usually in milliseconds or microseconds. Any network delay will have a significant impact on the transaction results and will bring a bad experience to customers. The traditional network architecture is difficult to meet the strict requirements of high-frequency trading for ultra-low latency, especially in the face of complex trading environments and changing market conditions. Data transmission bottlenecks, traffic congestion, and unpredictable jitter in the network may have a negative impact on transaction speed and accuracy. In addition, high-frequency trading traffic is generated in a short period of time, which is easy to impact other business traffic. Summary of the invention

[0003] In order to solve the above problems, the present invention provides a network optimization method for high-frequency trading, which comprises the following steps: A high-frequency transaction traffic identification unit is set in the OVS instance, and the high-frequency transaction traffic identification unit identifies the high-frequency transaction traffic and forwards the data packet to the DPDK data plane of the OVS instance, otherwise forwards it to the kernel data plane of the OVS instance; The DPDK data plane determines whether the data packet matches the flow table used by the DPDK data plane. If there is no match, it is forwarded to the kernel data plane. If there is still no matching flow table in the kernel data plane, the Packet-In event is triggered, and the control plane of the OVS instance processes the data packet and generates a new flow table entry; The flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency trading traffic identification unit and the flow table of the kernel data plane, and the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance.

[0004] Preferably, the high-frequency transaction flow identification unit identifies high-frequency flow, specifically: The high-frequency transaction traffic identification unit counts the average interval, average size, destination port, and transport layer protocol of data packets with the same source IP within a certain period of time. If the average interval, average size, destination port, and transport layer protocol all meet preset conditions, the data packets from the IP and all of which are TCP protocols are identified as high-frequency traffic data packets.

[0005] Preferably, the flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency transaction traffic identification unit and the flow table of the kernel data plane, specifically: Determine whether there is a flow table entry matching the high-frequency trading traffic in the kernel data plane. If so, load the matching flow table entry from the kernel data plane to the DPDK data plane and delete the matching flow table entry from the kernel data plane. The high-frequency transaction traffic identification unit continuously monitors whether the traffic is high-frequency transaction traffic. If it is no longer high-frequency transaction traffic, the flow table entry is loaded from the DPDK data plane to the kernel data plane and deleted from the DPDK data plane.

[0006] Preferably, the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance, specifically: The kernel data plane loads flow table entries that are no longer high-frequency traffic from the DPDK data plane, and loads newly generated flow table entries from the control plane of the OVS instance; The kernel data plane determines the flow table entry to be deleted from the kernel data plane according to the set idle timeout period.

[0007] Preferably, the OVS instance includes at least a control plane, a high-frequency transaction traffic identification unit, a DPDK data plane and a kernel data plane, wherein the control plane configures and manages the high-frequency transaction traffic identification unit, the DPDK data plane and the kernel data plane.

[0008] On the other hand, the present invention provides a network optimization system for high-frequency trading, the system comprising the following modules: A high-frequency transaction traffic identification unit, wherein the high-frequency transaction traffic identification unit is arranged in the data plane of the OVS instance, and the high-frequency transaction traffic identification unit identifies high-frequency transaction traffic and forwards the data packet to the DPDK data plane of the OVS instance, otherwise forwards it to the kernel data plane of the OVS instance; The data forwarding module is used by the DPDK data plane to determine whether the data packet matches the flow table used by the DPDK data plane. If there is no match, it is forwarded to the kernel data plane. If there is still no matching flow table in the kernel data plane, the Packet-In event is triggered, and the control plane of the OVS instance processes the data packet and generates a new flow table entry; The flow table update module is used to dynamically update the flow table of the DPDK data plane according to the statistical information of the high-frequency trading traffic identification unit and the flow table of the kernel data plane, and the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance.

[0009] Preferably, the high-frequency transaction flow identification unit identifies high-frequency flow, specifically: The high-frequency transaction traffic identification unit counts the average interval, average size, destination port, and transport layer protocol of data packets with the same source IP within a certain period of time. If the average interval, average size, destination port, and transport layer protocol all meet preset conditions, the data packets from the IP and all of which are TCP protocols are identified as high-frequency traffic data packets.

[0010] Preferably, the flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency transaction traffic identification unit and the flow table of the kernel data plane, specifically: Determine whether there is a flow table entry matching the high-frequency trading traffic in the kernel data plane. If so, load the matching flow table entry from the kernel data plane to the DPDK data plane and delete the matching flow table entry from the kernel data plane. The high-frequency transaction traffic identification unit continuously monitors whether the traffic is high-frequency transaction traffic. If it is no longer high-frequency transaction traffic, the flow table entry is loaded from the DPDK data plane to the kernel data plane and deleted from the DPDK data plane.

[0011] Preferably, the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance, specifically: The kernel data plane loads flow table entries that are no longer high-frequency traffic from the DPDK data plane, and loads newly generated flow table entries from the control plane of the OVS instance; The kernel data plane determines the flow table entry to be deleted from the kernel data plane according to the set idle timeout period.

[0012] Preferably, the OVS instance includes at least a control plane, a high-frequency transaction traffic identification unit, a DPDK data plane and a kernel data plane, wherein the control plane configures and manages the high-frequency transaction traffic identification unit, the DPDK data plane and the kernel data plane.

[0013] In view of the problem that high-frequency transactions have high latency requirements, the present invention sets a high-frequency transaction traffic identification unit in the data plane layer of OVS. If it is high-frequency transaction traffic, it is processed by the DPDK data plane of the data plane layer, otherwise it is handed over to the kernel data plane; and the DPDK data plane and kernel data plane flow tables of the present invention are updated in real time. The present invention not only improves the processing speed of high-frequency transactions, but also can reduce the impact of high-frequency transaction traffic on ordinary traffic by separating the DPDK data plane and the kernel data plane. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a flow chart of Embodiment 1; Figure 2 This is the architecture diagram of the OVS instance; Figure 3 A schematic diagram of the flow table flow between the control plane, kernel data plane, and DPDK data plane; Figure 4 This is a structural diagram of the second embodiment. DETAILED DESCRIPTION

[0015] The technical scheme in the embodiment of the present application will be described below in conjunction with the accompanying drawings in the embodiment of the present application. In the embodiment of the present application, "at least one" refers to one or more, and "multiple" refers to two or more. "And / or" describes the association relationship of the associated objects, indicating that there may be three relationships, for example, A and / or B, which can represent: A exists alone, A and B exist at the same time, and B exists alone, where A and B can be singular or plural. The character " / " generally indicates that the associated objects before and after are in an "or" relationship. "At least one of the following" or its similar expressions refers to any combination of these items, including any combination of single items or plural items. For example, at least one of a, b or c can represent: a, b, c, a and b, a and c, b and c or a, b and c, where a, b and c can be single or multiple. In addition, in the embodiment of the present application, the words "first", "second" and the like do not limit the quantity and order.

[0016] It should be noted that in the embodiments of the present application, words such as "exemplary" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary" or "for example" is intended to present related concepts in a specific way.

[0017] Embodiment 1 The present invention provides a network optimization method for high-frequency trading, such as Figure 1 As shown, the method comprises the following steps: S1, setting a high-frequency transaction traffic identification unit in the OVS instance, wherein the high-frequency transaction traffic identification unit identifies high-frequency transaction traffic and forwards the data packet to the DPDK data plane of the OVS instance, otherwise forwards it to the kernel data plane of the OVS instance; OVS (Open vSwitch) is an open source virtual switch, which is widely used in cloud computing and is the basis of Kube-OVN, providing network services for Pod, VM, etc. in cloud computing. In one embodiment, OVS includes a control plane and a data plane, wherein the control plane is also called OVS management program or ovs controller, which is responsible for managing the data plane, providing at least the configuration, management, processing of data packets that do not match the flow table of the kernel data plane, generating new flow table entries, etc. of the data plane, and the data plane completes operations such as forwarding of data packets. In one embodiment, the control plane of the OVS is an OpenFlow controller, that is, the OpenFlow controller is used as the control plane, of course, other programs can also be used as the control plane, and the control plane, data plane, etc. can also be implemented by programming.

[0018] A high-frequency transaction flow identification unit is added to the data plane of the present invention to identify high-frequency transaction flow. The so-called high-frequency transaction flow is a large number of data packets in a short period of time, and the size of the data packets is relatively small. For example, there will be a large number of data packets after the activities such as flash sales and rush purchases reach a certain time point. The high-frequency transaction flow identification unit identifies these data packets. If it is a data packet in the high-frequency transaction flow, the high-frequency transaction flow identification unit forwards the data packet to the DPDK data plane of the OVS instance, otherwise it is forwarded to the kernel data plane of the OVS instance. High-frequency transaction flow has high requirements for delay, and DPDK does not use the network stack of the Linux kernel, but directly processes it in user mode, which can reduce the overhead of context switching and data replication, and forwarding high-frequency transaction flow and ordinary flow distribution to the DPDK data plane and the kernel data plane can isolate the two, reducing the impact of high-frequency transaction flow on ordinary flow. Among them, the DPDK data plane is also called the DPDK mode, which means that the data packet does not pass through the kernel's network stack, but is a function or module or program that processes the data packet in the user space, and the kernel data plane is also called the kernel path mode, which means that the data packet is processed through the kernel's network stack. Function or module or program. Unless otherwise specified, DPDK data plane and DPDK mode have the same meaning, and kernel data plane and kernel path mode have the same meaning.

[0019] In one embodiment, an OVS instance includes at least a control plane and a data plane, wherein the data plane includes a high-frequency transaction traffic identification unit, a DPDK data plane, and a kernel data plane, that is, the high-frequency transaction traffic identification unit, the DPDK data plane, and the kernel data plane are located in the data plane, such as Figure 2 As shown in the figure, the control plane configures and manages the high-frequency trading traffic identification unit, the DPDK data plane, and the kernel data plane.

[0020] The high-frequency transaction traffic identification unit is mainly used to identify high-frequency transaction traffic. In one embodiment, the high-frequency transaction traffic identification unit identifies high-frequency traffic, specifically: The high-frequency transaction traffic identification unit counts the average interval, average size, destination port, and transport layer protocol of data packets with the same source IP within a certain period of time. If the average interval, average size, destination port, and transport layer protocol all meet preset conditions, the data packets from the IP and all of which are TCP protocols are identified as high-frequency traffic data packets.

[0021] Specifically, the libpcap library is used to capture the traffic of an OVS instance, i.e., a subnet, and extract relevant information therefrom, including but not limited to the source IP address, average packet interval, average packet size, destination port, and transport layer protocol of packets within a period of time, such as 5 seconds. If the average interval, average size, destination port, and transport layer protocol meet the preset conditions, the subsequent packets from the same IP and all using the TCP protocol are identified as high-frequency traffic packets. In one embodiment, the preset conditions are that the average interval is less than 10ms, the average size is less than 500 bytes, the destination port is 80, and the transport layer protocol is TCP.

[0022] The identification of high-frequency transaction traffic by the high-frequency transaction traffic identification unit is a dynamic process. If the preset conditions are not met, the data packet will be identified as ordinary traffic. For example, in the third 5 seconds, if the average interval is greater than 10ms, the subsequent data packets from the source IP will not be identified as high-frequency transaction traffic data packets.

[0023] The high-frequency trading traffic identification unit may also use other methods to identify high-frequency traffic, including but not limited to content-based deep packet inspection, machine learning, etc.

[0024] S2, the DPDK data plane determines whether the data packet matches the flow table used by the DPDK data plane. If there is no match, it is forwarded to the kernel data plane. If there is still no matching flow table in the kernel data plane, the Packet-In event is triggered, and the control plane of the OVS instance processes the data packet and generates a new flow table entry; In OVS, data packets are mainly processed by flow table matching. The flow table consists of multiple flow table entries. Each flow table entry includes a matching field and an action. If a data packet matches the matching field, the data will be operated on according to the execution of the action, such as forwarding, discarding, modifying, etc.

[0025] If the data packet matches the flow table of the DPDK data plane, the data packet is processed according to the operation specified in the flow table. If the DPDK data plane does not have a matching flow table, the DPDK data plane forwards the data packet to the kernel data plane. If the kernel data plane has a flow table entry that matches the data packet, the data packet is processed according to the operation of the flow table entry. If the kernel data plane does not have a flow table entry that matches the data packet, the kernel data plane triggers a Packet-In event, and then hands the data packet to the control plane of the OVS instance to process the data packet. The control plane processes the data packet according to the configuration or default method, and generates a new flow table entry. In the present invention, once the Packet-In event is triggered, the data packet will be encapsulated and sent to the control plane after operations such as encapsulation. The control plane processes the data packet and generates a new flow table entry, which is then sent to the kernel data plane.

[0026] S3, the flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency trading traffic identification unit and the flow table of the kernel data plane, and the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance.

[0027] During the operation of the OVS instance, the flow tables of the DPDK data plane and the kernel data plane are constantly updated and changed. The DPDK data plane has no flow table at the initial moment. As the high-frequency trading traffic identification unit forwards data to the DPDK data plane, the flow table of the DPDK data plane will be continuously updated.

[0028] In one embodiment, the flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency transaction traffic identification unit and the flow table of the kernel data plane, specifically: Determine whether there is a flow table entry matching the high-frequency trading traffic in the kernel data plane. If so, load the matching flow table entry from the kernel data plane to the DPDK data plane and delete the matching flow table entry from the kernel data plane. The high-frequency transaction traffic identification unit continuously monitors whether the traffic is high-frequency transaction traffic. If it is no longer high-frequency transaction traffic, the flow table entry is loaded from the DPDK data plane to the kernel data plane and deleted from the DPDK data plane.

[0029] The update of the flow table mainly includes the addition and deletion of flow table items. For the DPDK data plane, the addition of its flow table comes from the kernel data plane. Specifically, if a data packet is to be sent to the DPDK data plane, but the DPDK data plane does not have a matching flow table item, the DPDK data plane will load the flow table item that matches the data table from the kernel data plane. Due to the limited space for storing the flow table, after the DPDK data plane loads the flow table item, the flow table item of the kernel data plane will be deleted to save space. When it is identified that a certain high-frequency trading flow is no longer a high-frequency trading flow, the matching flow table item is sent from the DPDK data plane to the kernel data plane and deleted from the DPDK data plane. In addition, the management of flow table items in the DPDK data plane can also be combined with idle timeouts and hard timeouts, which will not be repeated here.

[0030] Similarly, the update of the flow table of the kernel data plane also includes the addition and deletion of flow table entries. In one embodiment, the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance, specifically: The kernel data plane loads flow table entries that are no longer high-frequency traffic from the DPDK data plane, and loads newly generated flow table entries from the control plane of the OVS instance; The kernel data plane determines the flow table entry to be deleted from the kernel data plane according to the set idle timeout period.

[0031] In addition to the configuration from the control plane, the flow table entries of the kernel data plane also come from new flow table entries generated by the control plane and flow table entries deleted from the DPDK data plane. The flow table entries of the kernel data plane are deleted based on the set idle timeout. On the other hand, if the DPDK data plane loads flow table entries from the kernel data plane, the flow table entries of the kernel data plane will also be deleted. Figure 3 The relationship between the flow table entries of the control plane, DPDK data plane, and kernel data plane is shown. It should be noted that steps S1, S2, and S3 here are only for the convenience of explanation and do not constitute a strict execution order. In particular, S2 and S3 are interspersed in execution. For example, after the first data table of high-frequency trading traffic in S2 is forwarded to the DPDK module, if the DPDK module does not have a matching flow table entry, it will be forwarded to the kernel data plane and trigger S3. Similarly, if the control plane generates a new flow table, it will trigger the kernel data plane to load new flow table entries from the control plane.

[0032] Embodiment 2 The present invention provides a network optimization system for high-frequency trading, such as Figure 4 , the system comprises the following modules: A high-frequency transaction traffic identification unit, wherein the high-frequency transaction traffic identification unit is arranged in the data plane of the OVS instance, and the high-frequency transaction traffic identification unit identifies high-frequency transaction traffic and forwards the data packet to the DPDK data plane of the OVS instance, otherwise forwards it to the kernel data plane of the OVS instance; The data forwarding module is used by the DPDK data plane to determine whether the data packet matches the flow table used by the DPDK data plane. If there is no match, it is forwarded to the kernel data plane. If there is still no matching flow table in the kernel data plane, the Packet-In event is triggered, and the control plane of the OVS instance processes the data packet and generates a new flow table entry; The flow table update module is used to dynamically update the flow table of the DPDK data plane according to the statistical information of the high-frequency trading traffic identification unit and the flow table of the kernel data plane, and the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance.

[0033] Preferably, the high-frequency transaction flow identification unit identifies high-frequency flow, specifically: The high-frequency transaction traffic identification unit counts the average interval, average size, destination port, and transport layer protocol of data packets with the same source IP within a certain period of time. If the average interval, average size, destination port, and transport layer protocol all meet preset conditions, the data packets from the IP and all of which are TCP protocols are identified as high-frequency traffic data packets.

[0034] Preferably, the flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency transaction traffic identification unit and the flow table of the kernel data plane, specifically: Determine whether there is a flow table entry matching the high-frequency trading traffic in the kernel data plane. If so, load the matching flow table entry from the kernel data plane to the DPDK data plane and delete the matching flow table entry from the kernel data plane. The high-frequency transaction traffic identification unit continuously monitors whether the traffic is high-frequency transaction traffic. If it is no longer high-frequency transaction traffic, the flow table entry is loaded from the DPDK data plane to the kernel data plane and deleted from the DPDK data plane.

[0035] Preferably, the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance, specifically: The kernel data plane loads flow table entries that are no longer high-frequency traffic from the DPDK data plane, and loads newly generated flow table entries from the control plane of the OVS instance; The kernel data plane determines the flow table entry to be deleted from the kernel data plane according to the set idle timeout period.

[0036] Preferably, the OVS instance includes at least a control plane, a high-frequency transaction traffic identification unit, a DPDK data plane and a kernel data plane, wherein the control plane configures and manages the high-frequency transaction traffic identification unit, the DPDK data plane and the kernel data plane.

[0037] Through the description of the above implementation methods, technical personnel in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional modules is used as an example. In actual applications, the above-mentioned functions can be assigned to different functional modules as needed, that is, the internal structure of the device can be divided into different functional modules to complete all or part of the functions described above.

[0038] In the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0039] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0040] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a readable storage medium. Based on this understanding, the part of the technical solution of the embodiment of the present application that contributes essentially or all or part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a device (which can be a single-chip microcomputer, chip, etc.) or a processor to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory, a random access memory, a magnetic disk or an optical disk.

[0041] Finally, it should be noted that the above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

Claims

1. A network optimization method for high-frequency trading, characterized in that: The method comprises the following steps: A high-frequency transaction traffic identification unit is set in the OVS instance, and the high-frequency transaction traffic identification unit identifies the high-frequency transaction traffic and forwards the data packet to the DPDK data plane of the OVS instance, otherwise forwards it to the kernel data plane of the OVS instance; The DPDK data plane determines whether the data packet matches the flow table used by the DPDK data plane. If there is no match, it is forwarded to the kernel data plane. If there is still no matching flow table in the kernel data plane, the Packet-In event is triggered, and the control plane of the OVS instance processes the data packet and generates a new flow table entry; The flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency trading traffic identification unit and the flow table of the kernel data plane, and the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance.

2. The method according to claim 1, characterized in that The high-frequency transaction flow identification unit identifies high-frequency flow, specifically: The high-frequency transaction traffic identification unit counts the average interval, average size, destination port, and transport layer protocol of data packets with the same source IP within a certain period of time. If the average interval, average size, destination port, and transport layer protocol all meet preset conditions, the data packets from the IP and all of which are TCP protocols are identified as high-frequency traffic data packets.

3. The method according to claim 1, characterized in that The flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency transaction traffic identification unit and the flow table of the kernel data plane, specifically: Determine whether there is a flow table entry matching the high-frequency trading traffic in the kernel data plane. If so, load the matching flow table entry from the kernel data plane to the DPDK data plane and delete the matching flow table entry from the kernel data plane. The high-frequency transaction traffic identification unit continuously monitors whether the traffic is high-frequency transaction traffic. If it is no longer high-frequency transaction traffic, the flow table entry is loaded from the DPDK data plane to the kernel data plane and deleted from the DPDK data plane.

4. The method according to claim 1, characterized in that The flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and control plane of the OVS instance, specifically: The kernel data plane loads flow table entries that are no longer high-frequency traffic from the DPDK data plane, and loads newly generated flow table entries from the control plane of the OVS instance; The kernel data plane determines the flow table entry to be deleted from the kernel data plane according to the set idle timeout period.

5. The method according to claim 1, characterized in that The OVS instance includes at least a control plane, a high-frequency transaction traffic identification unit, a DPDK data plane and a kernel data plane, wherein the control plane configures and manages the high-frequency transaction traffic identification unit, the DPDK data plane and the kernel data plane.

6. A network optimization system for high-frequency trading, characterized in that: The system includes the following modules: A high-frequency transaction traffic identification unit, wherein the high-frequency transaction traffic identification unit is arranged in the data plane of the OVS instance, and the high-frequency transaction traffic identification unit identifies high-frequency transaction traffic and forwards the data packet to the DPDK data plane of the OVS instance, otherwise forwards it to the kernel data plane of the OVS instance; The data forwarding module is used by the DPDK data plane to determine whether the data packet matches the flow table used by the DPDK data plane. If there is no match, it is forwarded to the kernel data plane. If there is still no matching flow table in the kernel data plane, the Packet-In event is triggered, and the control plane of the OVS instance processes the data packet and generates a new flow table entry; The flow table update module is used to dynamically update the flow table of the DPDK data plane according to the statistical information of the high-frequency trading traffic identification unit and the flow table of the kernel data plane, and the flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and the control plane of the OVS instance.

7. The system according to claim 6, characterized in that The high-frequency transaction flow identification unit identifies high-frequency flow, specifically: The high-frequency transaction traffic identification unit counts the average interval, average size, destination port, and transport layer protocol of data packets with the same source IP within a certain period of time. If the average interval, average size, destination port, and transport layer protocol all meet preset conditions, the data packets from the IP and all of which are TCP protocols are identified as high-frequency traffic data packets.

8. The system according to claim 6, characterized in that The flow table of the DPDK data plane is dynamically updated according to the statistical information of the high-frequency transaction traffic identification unit and the flow table of the kernel data plane, specifically: Determine whether there is a flow table entry matching the high-frequency trading traffic in the kernel data plane. If so, load the matching flow table entry from the kernel data plane to the DPDK data plane and delete the matching flow table entry from the kernel data plane. The high-frequency transaction traffic identification unit continuously monitors whether the traffic is high-frequency transaction traffic. If it is no longer high-frequency transaction traffic, the flow table entry is loaded from the DPDK data plane to the kernel data plane and deleted from the DPDK data plane.

9. The system according to claim 6, characterized in that The flow table of the kernel data plane is dynamically updated according to the flow tables of the DPDK data plane and control plane of the OVS instance, specifically: The kernel data plane loads flow table entries that are no longer high-frequency traffic from the DPDK data plane, and loads newly generated flow table entries from the control plane of the OVS instance; The kernel data plane determines the flow table entry to be deleted from the kernel data plane according to the set idle timeout period.

10. The system according to claim 6, characterized in that The OVS instance includes at least a control plane, a high-frequency transaction traffic identification unit, a DPDK data plane and a kernel data plane, wherein the control plane configures and manages the high-frequency transaction traffic identification unit, the DPDK data plane and the kernel data plane.

Images