CN116886550A - In-band network telemetry arrangement method and device in micro-service scene - Google Patents

Abstract

The application relates to an in-band network telemetry arrangement method and device in a micro-service scene, wherein the method comprises the following steps: receiving a telemetry data packet constructed by a network telemetry orchestration node, and forwarding the telemetry data packet to a next programmable switching node of the current programmable switching node; screening out qualified first telemetry data packets; determining according to the relation between a first equipment identification number in the equipment identification number list in the first telemetry data packet and the equipment identification number of the current programmable switching node, transmitting the first telemetry data packet to the next hop of the current programmable switching node until all micro services and all information of the programmable switching node are acquired to obtain a second telemetry data packet, or forwarding the first telemetry data packet to a network telemetry arrangement node for data processing; and forwarding the second telemetry data packet to the network telemetry orchestration node for data processing. The application can construct in-band network telemetry strategy aiming at the service call chain full link at one time.

Description

An in-band network telemetry orchestration method and device in a microservice scenario

Technical field

The present invention relates to the field of communication technology, and in particular to an in-band network telemetry orchestration method and device in a microservice scenario.

Background technique

In current cloud technology application scenarios, deployment models such as microservices and containerization are widely used. New businesses such as cloud computing and microservices that rely on high-performance network infrastructure have increasing demands on network latency, available bandwidth, or network availability, thus placing higher requirements on the performance of network measurement solutions.

In recent years, with the rise and development of new devices such as programmable switches and smart network cards, network programmability has been greatly improved. These advances have greatly expanded the design space of network measurement solutions. Among them, in-band network telemetry has brought a new direction to network measurement technology with its characteristics of real-time performance, accuracy, and no need for control plane participation. In-band network monitoring completes network status collection by sequentially inserting metadata (Measure metadata) into data packets through intermediate switching nodes in the path. Compared with traditional network measurement solutions, in-band measurement can achieve better understanding of network topology, network performance and network traffic. Fine-grained measurements. In specific application scenarios, it is often necessary to collect specific network traffic to monitor possible abnormal traffic. Based on the monitored traffic, fault location and fault elimination can be performed. At the same time, for specific business analysis scenarios for enterprise users, it is necessary to collect traffic with business-related attributes from designated network access ports.

In a microservice scenario, when the number of service calls in the service call chain is greater than 2, existing in-band network telemetry methods can only detect the service call relationships in the service call chain one by one, and cannot detect the entire service call chain at once.

Contents of the invention

In view of this, it is necessary to provide an in-band network telemetry orchestration method and device in a microservice scenario to achieve the purpose of completely constructing an in-band network telemetry strategy for the entire service call chain at one time.

In order to achieve the above objectives, the present invention provides an in-band network telemetry orchestration method in a microservice scenario, including:

Receive the telemetry data packet constructed by the network telemetry orchestration node, and forward the telemetry data packet to the next programmable switching node of the current programmable switching node. The telemetry data packet is transferred by the network telemetry orchestration node to the IP information of the current microservice. The device identification information of the current programmable switching node is written into the custom telemetry data packet constructed by the network telemetry orchestration node;

Filter out the telemetry data packets that meet the first preset condition to obtain the first telemetry data packet;

According to the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the device identification number of the current programmable switching node, the first telemetry data packet is sent to the current programmable switching node. The next hop of the node is until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet, or the first telemetry data packet is forwarded to the network telemetry orchestration node for data processing;

Forward the second telemetry data packet to a network telemetry orchestration node for data processing.

In some possible implementations, the telemetry data packet includes: IP header information, device identification number number, service instance IP list, device identification number list, in-band network telemetry data and payload information.

In some possible implementations, filtering out telemetry data packets that meet the first preset condition to obtain the first telemetry data packet includes:

Filter out the second telemetry data packet whose IP protocol number is the preset protocol number from the telemetry data packet;

The in-band telemetry data in the second telemetry data packet is updated to obtain a first telemetry data packet.

In some possible implementations, the first device identification number in the device identification number list in the first telemetry data packet is determined based on the relationship between the current programmable switching node device identification number and the first device identification number in the device identification number list. A telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected and a second telemetry data packet is obtained, including:

When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is not equal to the device identification number of the current programmable switching node, the first telemetry data packet is sent to the device identification number of the current programmable switching node. The next hop is until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet.

In some possible implementations, the first device identification number in the device identification number list in the first telemetry data packet is determined based on the relationship between the current programmable switching node device identification number and the first device identification number in the device identification number list. A telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected and a second telemetry data packet is obtained, including:

When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is equal to the current programmable switching node device identification number, and the device identification number in the device identification number list in the first telemetry data packet When the number is greater than 1, use the second IP in the service instance IP list in the first telemetry data packet as the destination IP in the IP header information in the first telemetry data packet;

Delete the second IP in the service instance IP list and the first device identification number in the device identification number list in the first telemetry data packet, and reduce the number of device identification numbers in the first telemetry data packet by 1 ;

The first telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet.

In some possible implementations, the first device identification number in the device identification number list in the first telemetry data packet is determined based on the relationship between the current programmable switching node device identification number and the first device identification number in the device identification number list. A telemetry data packet is forwarded to the network telemetry orchestration node for data processing, including:

When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is equal to the current programmable switching node device identification number, and the device identification number in the device identification number list in the first telemetry data packet When the number is not greater than 1, the first telemetry data packet is forwarded to the network telemetry orchestration node for data processing.

In some possible implementations, the programmable switching node is a switch.

On the other hand, the present invention also provides an in-band network telemetry orchestration device in a microservice scenario, including:

A data receiving module, configured to receive a custom telemetry data packet constructed by a network telemetry orchestration node, and forward the custom telemetry data packet to the next programmable switching node of the current programmable switching node, where the custom telemetry data packet Obtained by the network telemetry orchestration node writing the IP information of the current microservice and the device identification information of the current programmable switching node into a custom telemetry data packet constructed by the network telemetry orchestration node;

The first data judgment module is used to filter out custom telemetry data packets that meet the first preset condition to obtain the first telemetry data packet;

The second data judgment module is configured to determine, based on the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the current programmable switching node device identification number, the first telemetry The data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet, or the first telemetry data packet is forwarded to the network Telemetry orchestration nodes perform data processing;

A data forwarding module, configured to forward the second telemetry data packet to a network telemetry orchestration node for data processing.

The beneficial effects of adopting the above embodiments are: the invention provides an in-band network telemetry orchestration method in a microservice scenario, receives telemetry data packets constructed by network telemetry orchestration nodes, and forwards the telemetry data packets to the current programmable The next programmable switching node of the switching node filters the telemetry data packets to obtain the first telemetry data packet. According to the first device identification number in the device identification number list in the first telemetry data packet and the current programmable switching node device identification The relationship between the numbers is determined and the first telemetry data packet is directly forwarded to the network telemetry orchestration node, or continues to be forwarded to the next hop until the information of all microservices and all programmable switching nodes is collected and the second telemetry data packet is obtained. Forwarding the second telemetry data packet to the network telemetry orchestration node. The invention writes hierarchical service chain IP information and programmable switching node device identification information into customized telemetry data packets, thereby detecting the entire service calling chain at one time.

Description of the drawings

Figure 1 is a method flow chart of an in-band network telemetry orchestration method in a microservice scenario provided by the present invention;

Figure 2 is a schematic diagram of the service grid management node and programmable switching node invocation of an in-band network telemetry orchestration method in a microservice scenario provided by the present invention;

Figure 3 is a schematic diagram of the custom telemetry data packet processing flow of an in-band network telemetry orchestration method in a microservice scenario provided by the present invention;

Figure 4 is a schematic structural diagram of an embodiment of an in-band network telemetry orchestration device in a microservice scenario provided by the present invention.

Detailed ways

The preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The drawings constitute a part of this application and are used together with the embodiments of the present invention to illustrate the principles of the present invention, but are not intended to limit the scope of the present invention.

Figure 1 is a schematic flow chart of an embodiment of an in-band network telemetry orchestration method in a microservice scenario provided by the present invention. As shown in Figure 1, an in-band network telemetry orchestration method in a microservice scenario includes:

S101. Receive the telemetry data packet constructed by the network telemetry orchestration node, and forward the telemetry data packet to the next programmable switching node of the current programmable switching node. The telemetry data packet is forwarded by the network telemetry orchestration node to the current microservice. The IP information and the device identification information of the current programmable switching node are written into the custom telemetry data packet constructed by the network telemetry orchestration node;

S102. Filter out the telemetry data packets that meet the first preset condition to obtain the first telemetry data packet;

S103. According to the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the current programmable switching node device identification number, send the first telemetry data packet to the currently programmable switching node device identification number. Program the next hop of the switching node until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet, or forward the first telemetry data packet to the network telemetry orchestration node for data processing ;

S104. Forward the second telemetry data packet to the network telemetry orchestration node for data processing.

Compared with the existing technology, this embodiment provides an in-band network telemetry orchestration method in a microservice scenario.

Receive the telemetry data packet constructed by the network telemetry orchestration node, forward the telemetry data packet to the next programmable switching node of the current programmable switching node, filter the telemetry data packet to obtain the first telemetry data packet, and obtain the first telemetry data packet according to the first telemetry data packet. The relationship between the first device identification number in the device identification number list in the data packet and the current programmable switching node device identification number determines whether the first telemetry data packet is forwarded directly to the network telemetry orchestration node, or continues to be forwarded to the next hop. Until the information of all microservices and all programmable switching nodes is collected and the second telemetry data packet is obtained, the second telemetry data packet is forwarded to the network telemetry orchestration node. The invention writes hierarchical service chain IP information and programmable switching node device identification information into customized telemetry data packets, thereby detecting the entire service calling chain at one time.

It should be noted that the execution subject of this embodiment is a programmable switching node.

It should be noted that the service grid management node refers to a programmable switching device with in-band network telemetry capability in the data plane. In this embodiment, the service grid management node is a microservice management device on a microservice. The programmable switching node refers to a programmable switching device with in-band network telemetry capability in the data plane. In this embodiment, the programmable switching node is a switch, specifically a P4 switch. The network telemetry orchestration node refers to a device in the control plane that is responsible for in-band network telemetry task planning and analysis of data plane feedback data. In this embodiment, the network telemetry orchestration node is a server.

In some embodiments of the present invention, the telemetry data packet includes: IP header information, device identification number number, service instance IP list, device identification number list, in-band network telemetry data and load information. The number of IPs in the service instance IP list is the same as the number of device identification numbers in the device identification number list.

In the specific embodiment of the present invention, three service instances, namely service grid management nodes SvcA, SvcB, and SvcC, and five programmable switching nodes SW1, SW2, SW3, SW4, and SW5 are taken as an example, in which SvcA is directly connected to SW1. , SvcB is directly connected to SW3, SvcC is directly connected to SW5, SW2 is connected to SW1 and SW3, and SW4 is connected to SW3 and SW5. Figure 2 is a service grid of an in-band network telemetry orchestration method in a microservice scenario provided by the present invention. The schematic diagram of the call between the management node and the programmable switching node is shown in Figure 2. When there is a call chain SvcA->SvcB->SvcC, the service instance IP list is the IP of SvcA, the IP of SvcB, the IP of SvcC, and the service instance The list of connected programmable switching node device identification numbers SWID is the SWID of SW1, the SWID of SW3, and the SWID of SW3. The in-band network telemetry data in the custom telemetry data package adopts the same method as the existing in-band network telemetry. When the in-band network telemetry source is used, the in-band network telemetry is initialized and the metadata of the switching node is added. Each time it passes through a Programmable switching nodes all add metadata of switching nodes.

In some embodiments of the present invention, filtering out telemetry data packets that meet the first preset condition to obtain the first telemetry data packet includes:

Filter out the second telemetry data packet whose IP protocol number is the preset protocol number from the telemetry data packet;

The in-band telemetry data in the second telemetry data packet is updated to obtain a first telemetry data packet.

In some embodiments of the present invention, the determination is made based on the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the current programmable switching node device identification number. The first telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected and the second telemetry data packet is obtained, including:

When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is not equal to the device identification number of the current programmable switching node, the first telemetry data packet is sent to the device identification number of the current programmable switching node. The next hop is until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet.

In some embodiments of the present invention, the determination is made based on the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the current programmable switching node device identification number. The first telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected and the second telemetry data packet is obtained, including:

When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is equal to the current programmable switching node device identification number, and the device identification number in the device identification number list in the first telemetry data packet When the number is greater than 1, use the second IP in the service instance IP list in the first telemetry data packet as the destination IP in the IP header information in the first telemetry data packet;

Delete the second IP in the service instance IP list and the first device identification number in the device identification number list in the first telemetry data packet, and reduce the number of device identification numbers in the first telemetry data packet by 1 ;

The first telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet.

In some embodiments of the present invention, the determination is made based on the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the current programmable switching node device identification number. The first telemetry data packet is forwarded to the network telemetry orchestration node for data processing, including:

When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is equal to the current programmable switching node device identification number, and the device identification number in the device identification number list in the first telemetry data packet When the number is not greater than 1, the first telemetry data packet is forwarded to the network telemetry orchestration node for data processing.

In some embodiments of the invention, the programmable switching node is a switch.

In a specific embodiment of the present invention, Figure 3 is a schematic diagram of the custom telemetry data packet processing flow of an in-band network telemetry orchestration method in a microservice scenario provided by the present invention:

Step 1: Identify the custom telemetry data packet with IP protocol number 200, initialize the in-band network telemetry information, add metadata of the switching node, and obtain the first telemetry data packet;

Step 2: Further process the first telemetry data packet. The processing process is divided into three situations:

Case 1: When the device identification number list in the first telemetry data packet, that is, the first SWID in the SWID list is equal to the local SWID and the device identification number is greater than 1, fill in the second IP in the service instance IP list into the first telemetry For the destination IP of the data packet, delete the second IP in the service instance IP list, delete the first SWID in the SWID list, decrement the device identification number by one, and forward it according to the routing table/switching table.

Case 2: When the first SWID in the SWID list in the first telemetry packet is not equal to the local SWID, it is forwarded according to the routing table/switching table.

Case 3: When the first SWID in the SWID list in the first telemetry packet is equal to the local SWID and the device identification number is equal to 1, then the in-band network telemetry information is aggregated and reported to the network telemetry orchestration node.

As shown in Figure 4, an embodiment of the present invention also provides an in-band network telemetry orchestration device in a microservice scenario. An in-band network telemetry orchestration device 400 in a microservice scenario includes:

The data receiving module 401 is used to receive the custom telemetry data packet constructed by the network telemetry orchestration node, and forward the custom telemetry data packet to the next programmable switching node of the current programmable switching node. The package is obtained by the network telemetry orchestration node writing the IP information of the current microservice and the device identification information of the current programmable switching node into a custom telemetry data packet constructed by the network telemetry orchestration node;

The first data judgment module 402 is used to filter out custom telemetry data packets that meet the first preset condition to obtain the first telemetry data packet;

The second data judgment module 403 is configured to determine, based on the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the current programmable switching node device identification number, the first device identification number. The telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet, or the first telemetry data packet is forwarded to Network telemetry orchestration nodes perform data processing;

The data forwarding module 404 is used to forward the second telemetry data packet to the network telemetry orchestration node for data processing.

Those skilled in the art can understand that all or part of the process of implementing the method of the above embodiments can be completed by instructing relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. Wherein, the computer-readable storage medium is a magnetic disk, an optical disk, a read-only memory or a random access memory, etc.

The above are only preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person familiar with the technical field can easily think of changes or modifications within the technical scope disclosed in the present invention. All substitutions are within the scope of the present invention.

Claims (10)

1. An in-band network telemetry orchestration method in a microservice scenario, which is characterized by including: Receive the telemetry data packet constructed by the network telemetry orchestration node, and forward the telemetry data packet to the next programmable switching node of the current programmable switching node. The telemetry data packet is transferred by the network telemetry orchestration node to the IP information of the current microservice. The device identification information of the current programmable switching node is written into the custom telemetry data packet constructed by the network telemetry orchestration node; Filter out the telemetry data packets that meet the first preset condition to obtain the first telemetry data packet; According to the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the device identification number of the current programmable switching node, the first telemetry data packet is sent to the current programmable switching node. The next hop of the node is until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet, or the first telemetry data packet is forwarded to the network telemetry orchestration node for data processing; Forward the second telemetry data packet to a network telemetry orchestration node for data processing. 2. An in-band network telemetry orchestration method in a microservice scenario according to claim 1, characterized in that the telemetry data packet includes: IP header information, device identification number number, service instance IP list, device List of identification numbers, in-band network telemetry data, and payload information. 3. An in-band network telemetry orchestration method in a microservice scenario according to claim 2, characterized in that the telemetry data packets that meet the first preset condition are filtered out to obtain the first telemetry data packet, including : Filter out the second telemetry data packet whose IP protocol number is the preset protocol number from the telemetry data packet; The in-band telemetry data in the second telemetry data packet is updated to obtain a first telemetry data packet. 4. An in-band network telemetry orchestration method in a microservice scenario according to claim 1, characterized in that: according to the first device identification number in the device identification number list in the first telemetry data packet The relationship with the device identification number of the current programmable switching node is determined, and the first telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected. After completion, the second telemetry data packet is obtained, including: When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is not equal to the device identification number of the current programmable switching node, the first telemetry data packet is sent to the device identification number of the current programmable switching node. The next hop is until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet. 5. An in-band network telemetry orchestration method in a microservice scenario according to claim 1, characterized in that: according to the first device identification number in the device identification number list in the first telemetry data packet The relationship with the device identification number of the current programmable switching node is determined, and the first telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected. After completion, the second telemetry data packet is obtained, including: When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is equal to the current programmable switching node device identification number, and the device identification number in the device identification number list in the first telemetry data packet When the number is greater than 1, use the second IP in the service instance IP list in the first telemetry data packet as the destination IP in the IP header information in the first telemetry data packet; Delete the second IP in the service instance IP list and the first device identification number in the device identification number list in the first telemetry data packet, and reduce the number of device identification numbers in the first telemetry data packet by 1 ; The first telemetry data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet. 6. An in-band network telemetry orchestration method in a microservice scenario according to claim 1, characterized in that: according to the first device identification number in the device identification number list in the first telemetry data packet The relationship with the current programmable switching node device identification number is determined, and the first telemetry data packet is forwarded to the network telemetry orchestration node for data processing, including: When it is determined that the first device identification number in the device identification number list in the first telemetry data packet is equal to the current programmable switching node device identification number, and the device identification number in the device identification number list in the first telemetry data packet When the number is not greater than 1, the first telemetry data packet is forwarded to the network telemetry orchestration node for data processing. 7. An in-band network telemetry orchestration method in a microservice scenario according to claim 1, wherein the programmable switching node is a switch. 8. An in-band network telemetry orchestration device for microservice scenarios, which is characterized by including: A data receiving module, configured to receive a custom telemetry data packet constructed by a network telemetry orchestration node, and forward the custom telemetry data packet to the next programmable switching node of the current programmable switching node, where the custom telemetry data packet Obtained by the network telemetry orchestration node writing the IP information of the current microservice and the device identification information of the current programmable switching node into a custom telemetry data packet constructed by the network telemetry orchestration node; The first data judgment module is used to filter out custom telemetry data packets that meet the first preset condition to obtain the first telemetry data packet; The second data judgment module is configured to determine, based on the relationship between the first device identification number in the device identification number list in the first telemetry data packet and the current programmable switching node device identification number, the first telemetry The data packet is sent to the next hop of the current programmable switching node until the information of all microservices and all programmable switching nodes is collected to obtain the second telemetry data packet, or the first telemetry data packet is forwarded to the network Telemetry orchestration nodes perform data processing; A data forwarding module, configured to forward the second telemetry data packet to a network telemetry orchestration node for data processing. 9. An in-band network telemetry orchestration device in a microservice scenario according to claim 8, wherein the telemetry data packet includes: IP header information, device identification number number, service instance IP list, device List of identification numbers, in-band network telemetry data, and payload information. 10. An in-band network telemetry orchestration device in a microservice scenario according to claim 9, characterized in that the telemetry data packets that meet the first preset condition are filtered out to obtain the first telemetry data packet, including : Filter out the second telemetry data packet whose IP protocol number is the preset protocol number from the telemetry data packet; The in-band telemetry data in the second telemetry data packet is updated to obtain a first telemetry data packet.