Csig telemetry method based on hop count triggering, network forwarding device and system

Abstract

This invention belongs to the field of network communication technology and provides a CSIG telemetry method, network forwarding device, and system based on hop count triggering. It aims to address the technical shortcomings of existing IFA in-band measurement schemes, such as excessive packet expansion and overhead, and the inability of native CSIG aggregation schemes to achieve full-path panoramic data acquisition. This invention deploys specific processing logic in the network forwarding device to identify telemetry packets carrying CSIG tags with a new protocol identifier (NEW TPID), decrements the acquisition trigger counter (TTC), and triggers nodes to collect telemetry data and write it to the tag based on the TTC value. This invention can be applied to path node status acquisition and congestion monitoring in data center networks and campus networks, balancing lightweight characteristics with refined operation and maintenance requirements.

Description

Technical Field

[0001] This invention belongs to the field of network communication technology, specifically involving in-band network telemetry (INT) technology, which is particularly suitable for lightweight network telemetry scenarios based on the CSIG (Congestion Signal) protocol, and can be applied to path node status collection and congestion monitoring in data center networks and campus networks. Background Technology

[0002] With the continuous expansion of communication network scale and the increasing diversification of service types, the collection of network node operational status data has become a core foundation for network operation and maintenance management, performance monitoring and optimization, and network fault diagnosis and location. Currently, mainstream network data collection schemes are mainly divided into two categories: centralized network management polling (Pull) schemes and distributed proxy reporting (Push) schemes. Both of these schemes rely on the network control plane or out-of-band management channels to achieve data collection. The forwarding paths and timing of collected data and service data are not synchronized, generally resulting in poor real-time measurement performance, complex configuration and maintenance processes, and limited collection capabilities and scalability after network expansion.

[0003] To address the inherent limitations of out-of-band data acquisition schemes, in-band network telemetry technology has emerged. This technology embeds measurement commands and measurement metadata into service packets, enabling real-time measurement along with the flow of service data in the forwarding plane. This achieves "service traffic is measurement traffic," significantly improving the real-time performance and synchronization of network measurements, and has become an important technological development direction in the field of network measurement. Among them, Inband Flow Analyzer (IFA) is a typical in-band measurement architecture. In the IFA scheme, probe packets or service packets carry complete measurement commands, requiring each hop device on the forwarding path to forcibly append local node operating status information (such as queue delay, congestion flags, port information, etc.) to the measurement field of the packet. Although this scheme can obtain the full node status information of the entire forwarding path, as the number of packet forwarding hops increases, the amount of measurement data carried by the packet will continue to grow, which is very likely to exceed the limit of the network's maximum transmission unit (MTU), thus causing packet fragmentation. This not only significantly increases network bandwidth overhead, but also increases the forwarding processing burden of nodes along the way. In large-scale networking or long forwarding path application scenarios, there are obvious difficulties in deployment and promotion.

[0004] As an evolution of IFA technology, Congestion Signaling (CSIG) aims to design more flexible and scalable in-band measurement labels to address the packet bloat problem inherent in IFA schemes. However, existing CSIG implementations, in their pursuit of extreme simplification of measurement labels and reduction of bandwidth overhead, have extended to an "aggregated" measurement mode. For example, when measuring metrics such as maximum latency and minimum throughput, the measurement label only retains the worst-case metric value (e.g., maximum latency, minimum throughput) and its corresponding node identifier among the currently traversed nodes, without "accumulating" the independent measurement information of each hop on the path. While this method fundamentally solves the problem of measurement labels continuously bloating with the number of forwarding hops, it loses detailed measurement data for nodes other than bottleneck nodes on the forwarding path. It can only reflect the extreme bottleneck state in the path, making it impossible to perform precise point-to-point data collection for specific nodes. This results in poor adaptability to operational scenarios and makes it difficult to meet the needs of refined fault location and congestion investigation.

[0005] Furthermore, the tag field of the native CSIG protocol has a single function and is not reusable. It lacks a flexible triggering mechanism for data collection, has poor protocol compatibility, does not support a smooth transition between old and new devices, and lacks a sound exception handling and data recycling mechanism. This results in incomplete and uncontrollable telemetry data collection, which further limits its deployment and application in actual networks. Summary of the Invention

[0006] To address the aforementioned deficiencies in the prior art, this application provides a CSIG telemetry method, network forwarding device, and system based on hop count triggering, aiming to solve the following specific technical problems: 1) Solve the problem of excessive message expansion, bandwidth and processing overhead in the IFA in-band measurement scheme, which makes it unsuitable for long-path, large-scale networks; 2) Address the issue that the native CSIG aggregation solution can only collect data from bottleneck nodes and cannot achieve full-path panoramic data collection or fixed-point data collection; 3) Solve the problems of low utilization of native CSIG tag fields and inflexible triggering collection mechanisms.

[0007] To address the aforementioned technical problems, based on one aspect of the present invention, a network forwarding device is provided, including a processor and a memory, wherein the memory stores executable instructions, and the processor executes the executable instructions to implement the following steps: Receive telemetry messages carrying congestion signal tags, i.e., CSIG tags, and identify the new protocol identifier, i.e., NEW TPID, in the CSIG tags; If the NEW TPID is detected, read the trigger flag bit in the CSIG tag and collect the trigger counter TTC value; If the trigger flag is in a counting state, the TTC value is decremented by 1. If the TTC value after decrementing by 1 is 0, the trigger flag is set to the triggered state, telemetry data is collected, the telemetry data is written into the CSIG tag (signal value and location metadata fields), and the telemetry message is forwarded. If the TTC value is greater than 0 after being decremented by 1, the trigger flag is kept in the counting state and the TTC value in the CSIG tag is updated; the telemetry data includes the signal value corresponding to the signal type in the CSIG tag and the current location metadata; If the trigger flag is in the triggered state, transparent forwarding is performed on the telemetry message.

[0008] Furthermore, the NEW TPID adopts a segmented format, including a general CSIG TPID identifier segment and a tag version number segment. When the processor executes the executable instruction, the step of identifying the NEW TPID specifically includes: first performing a full-bit match on the NEW TPID; if the match is successful, it is determined that the NEW TPID has been identified; if the match is unsuccessful, then the general CSIG TPID identifier segment is matched; if the general CSIG TPID identifier segment is successfully matched, it is determined that the general CSIG TPID identifier has been identified. The network forwarding device processes the CSIG tag according to the general CSIG tag processing method (i.e., according to the native CSIG protocol).

[0009] Furthermore, the CSIG tag includes a compact format tag and an extended format tag, which can adopt two parallel field reuse schemes. When the processor executes the executable instruction, it selects one of the schemes to perform the corresponding operation according to a preset global unified configuration.

[0010] The first field reuse scheme is as follows: The compact format tag's bit structure is defined as follows: 16-bit NEW TPID, 3-bit signal type field, 1-bit trigger flag (original R field), 7-bit dual-state multiplexing field (TTC / LM multiplexing), and 5-bit signal value field (original S field); wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is in the "counting state" value (state value is 1); and serves as the location metadata LM field when the trigger flag is in the "triggered state" value (state value is 0). The extended format tag's bit structure is defined as follows: 16-bit NEW TPID, 16-bit dual-state multiplexing field (TTC / LM multiplexing), 4-bit signal type field, 20-bit signal value field, 1-bit trigger flag (1 bit in the original R field), and 7-bit spare extended field (the remaining 7 bits in the original R field). The dual-state multiplexing field serves as a TTC counter when the trigger flag is in the "counting state" value (state value 1); and as a location metadata LM field when the trigger flag is in the "triggered state" value (state value 0).

[0011] The second field reuse scheme is: The compact format tag's bit structure is defined as follows: 16-bit NEW TPID, 3-bit signal type field (formerly T field), 1-bit trigger flag (formerly R field), 7-bit location metadata field (formerly LM field), and 5-bit dual-state multiplexing field (TTC / S multiplexing); wherein, the dual-state multiplexing field acts as a TTC counter when the trigger flag is in the "counting state" value (state value 1); and acts as a signal value field when the trigger flag is in the "triggered state" value (state value 0). The extended format tag's bit structure is defined as follows: 16-bit NEW TPID, 16-bit location metadata field (original LM field), 4-bit signal type field (original T field), 20-bit dual-state multiplexing field (TTC / S multiplexing), 1-bit trigger flag (1 bit in the original R field), and 7-bit spare extended field (the remaining 7 bits in the original R field). The dual-state multiplexing field acts as a TTC counter when the trigger flag is in the "counting state" value (state value 1), and as a signal value field when the trigger flag is in the "triggered state" value (state value 0).

[0012] Furthermore, the telemetry messages include service messages and dedicated detection messages; If the telemetry message is a service message, then after telemetry data collection, it also includes: Determine the preset telemetry data collection mode. If it is the direct reporting mode, the telemetry data is directly reported to the management node through the out-of-band management channel. If it is the normal reporting mode, the steps of writing the telemetry data into the CSIG tag and forwarding the telemetry message are executed so that the destination device can report. If the telemetry message is a dedicated probe message, after the network forwarding device collects telemetry data, it constructs a CSIG feedback message, copies the telemetry data to the CSIG tag of the feedback message, and then sends it back to the source device or management node, or directly reports it to the management node through the out-of-band management network.

[0013] Based on another aspect of the embodiments of the present invention, the present invention also provides a CSIG telemetry method based on hop count triggering, applied to network forwarding devices, comprising the following steps: Receive telemetry messages carrying congestion signal tags, i.e., CSIG tags, and identify the new protocol identifier, i.e., NEW TPID, in the CSIG tags; if the NEW TPID is identified, read the trigger flag bit and collect the trigger counter (TTC) value in the CSIG tags; If the trigger flag is in a counting state, the TTC value is decremented by 1. If the TTC value after decrementing is 0, the trigger flag is set to the triggered state, telemetry data is collected, the telemetry data is written to the CSIG tag (signal value and location metadata fields), and the telemetry message is forwarded. If the TTC value after decrementing is greater than 0, the trigger flag is kept in a counting state, and the TTC value in the CSIG tag is updated. The telemetry data includes the signal value corresponding to the signal type in the CSIG tag and the current location metadata. If the trigger flag is in the triggered state, transparent forwarding is performed on the telemetry message.

[0014] Furthermore, the CSIG tag includes a compact format tag and an extended format tag, and can adopt two parallel field multiplexing schemes. The network forwarding device selects one of the schemes to perform the corresponding operation according to the preset global unified configuration. The specific bit definitions of the two field multiplexing schemes are consistent with the field multiplexing schemes in the network forwarding device mentioned above.

[0015] Furthermore, the telemetry message includes service messages and dedicated probe messages; if the telemetry message is a service message, after collecting telemetry data, the process further includes: determining a preset telemetry data recovery mode; if it is a direct reporting mode, the telemetry data is directly reported to the management node through the out-of-band management channel; if it is a regular reporting mode, the process involves writing the telemetry data into a CSIG tag and forwarding the telemetry message so that the destination device can report; if the telemetry message is a dedicated probe message, after collecting telemetry data, the network forwarding device further includes: constructing a CSIG feedback message, copying the telemetry data to the CSIG tag of the feedback message and sending it back to the source device or management node, or directly reporting it to the management node through the out-of-band management network.

[0016] Based on another aspect of the embodiments of the present invention, the present invention also provides a hop-count-triggered CSIG telemetry system, including a management node, a source device, a network forwarding device, and a destination device; the management node is used to detect the total number of hops of the probe path from the source device to the destination device through the Traceroute mechanism of the TCP / IP protocol, and send a telemetry configuration command to the source device; the source device is used to construct a telemetry message carrying a CSIG tag according to the telemetry configuration command, set the initial value of TTC in the CSIG tag, encapsulate it with a fixed 5-tuple, and send it to the network forwarding device; the network forwarding device is used to execute the above-described hop-count-triggered CSIG telemetry method; the destination device is used to receive the telemetry message, extract the telemetry data in the CSIG tag, and report it to the management node through an out-of-band management channel.

[0017] Furthermore, the telemetry message includes service messages and dedicated probe messages; if the telemetry message is a service message, the destination device extracts telemetry data from the telemetry message and reports it to the management node through the out-of-band management channel; if the telemetry message is a dedicated probe message, after triggering the collection, the network forwarding device constructs a CSIG feedback message, copies the telemetry data to the CSIG tag of the feedback message and sends it back to the source device or management node, or directly reports it to the management node through the out-of-band management network.

[0018] Furthermore, the management node is also used to: if a route switch is detected that causes a path change, re-obtain the new total path hop count through the Traceroute mechanism and send an updated telemetry configuration command to the source device; receive telemetry data reported by the destination device or network forwarding device, bind the telemetry data to the path nodes based on the total path hop count, and reconstruct a complete path view.

[0019] The beneficial effects of this invention are derived reasonably from the above technical solutions, as follows: 1) Solving the problem of IFA message bloat: This invention maintains a fixed length of CSIG label, does not write data hop by hop, avoids MTU fragmentation, significantly reduces bandwidth and processing overhead, and can be adapted to long-path, large-scale network deployments; 2) Overcome the shortcomings of native CSIG aggregation: Abandon the extreme value replacement logic of native CSIG, and realize fixed-point collection, path segment or full path panoramic collection through the TTC triggering mechanism. The operation and maintenance can obtain the running status of all nodes, which greatly improves the accuracy of fault location. 3) Improve tag utilization: Design two parallel schemes: TTC reuse of LM field and TTC reuse of S field, to expand tag functionality without increasing tag length and to be compatible with native CSIG protocol format; 4) Easy deployment both backwards and forwards: NEW TPID adopts a mask segmentation design, which can achieve a smooth transition between new and old devices and can be deployed without upgrading all network devices, reducing deployment costs; 5) Construct a closed-loop recycling mechanism: Adapt to two mainstream scenarios, passive telemetry (service message carrying labels) and active telemetry (dedicated probe message carrying labels), and provide multiple data recycling methods to ensure that telemetry data is not lost; 6) Ensure accurate hop count matching: Relying on the Traceroute mechanism of the TCP / IP protocol to detect the total number of hops in the path, ensuring that telemetry data and path nodes correspond accurately, and improving the reliability and availability of telemetry data.

[0020] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this specification. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the hardware structure of the network forwarding device in an embodiment of the present invention; Figure 2 This is a flowchart illustrating the CSIG telemetry method based on hop count triggering in an embodiment of the present invention; Figure 3 This is a schematic diagram illustrating the compact format tag bit definition of the first field reuse scheme in an embodiment of the present invention; Figure 4 This is a schematic diagram illustrating the extended format tag bit definition of the first field multiplexing scheme in this embodiment of the invention; Figure 5 This is a schematic diagram illustrating the compact format tag bit definition of the second field reuse scheme in this embodiment of the invention; Figure 6 This is a schematic diagram illustrating the extended format tag bit definition of the second field multiplexing scheme in this embodiment of the invention; Figure 7 This is a schematic diagram of the CSIG telemetry system based on hop count triggering in an embodiment of the present invention.

[0022] Note: The accompanying drawings are for illustrative purposes only and are used to help explain the technical solution of the present invention. They do not constitute a limitation on the scope of protection of the present invention. All references in the drawings are consistent with the text descriptions in the specification to ensure logical correspondence. Detailed Implementation

[0023] The present invention will be further illustrated below with reference to specific embodiments. It should be understood that these embodiments are for illustrative purposes only and are not intended to limit the scope of the invention. Furthermore, it should be understood that after reading the teachings of this invention, those skilled in the art can make various alterations or modifications to the invention, and these equivalent forms also fall within the scope defined by the appended claims.

[0024] Example 1 like Figure 1 As shown, this embodiment provides a network forwarding device 100, including a processor 101, a memory 102, and a network interface 103. The processor, memory, and network interface are electrically connected via a bus. The memory stores executable instructions, and the processor executes these instructions to perform all operations related to telemetry message processing, CSIG tag identification, TTC counting, telemetry data acquisition, and reporting. The network interface is used to send and receive messages. The processor can be a general-purpose processor, such as a central processing unit (CPU), microprocessor, or network processor. The memory can be random access memory (RAM), read-only memory (ROM), etc., used to store executable instructions related to telemetry message processing, CSIG tag identification, TTC counting, telemetry data acquisition, and reporting, as well as temporarily storing data generated during telemetry (such as TTC values, telemetry data, trigger flag status, etc.).

[0025] Specifically, the processor executes executable instructions to perform the following steps: The first step is to receive telemetry messages carrying the CSIG tag. The network forwarding device receives telemetry messages from the source device or the upstream network forwarding device through its own message receiving interface (such as an Ethernet interface). The telemetry message carries the CSIG tag, which is used to identify telemetry-related protocol information, trigger information, and data information. It includes core fields such as the New Protocol Identifier (NEW TPID), trigger flag, and acquisition trigger counter (TTC) value.

[0026] The new protocol identifier (NEW Tag Protocol ID, NEW TPID) is used to identify the new CSIG tag that supports the hop count-triggered acquisition mechanism provided by this invention. The trigger flag (F) is used to identify the state of triggering telemetry data acquisition, including "counting state" and "triggered state". The acquisition trigger counter (Time To Collect, TTC) is used to record the number of hops between the current position and the acquisition position, and the telemetry data acquisition position is controlled by preset TTC values.

[0027] The second step is to identify the NEW TPID in the CSIG label.

[0028] In a preferred embodiment of the present invention, the NEW TPID adopts a segmented format, including a general CSIG TPID identifier segment and a tag version number segment. The general CSIG TPID identifier segment is used to achieve compatibility with the native CSIG protocol, and the tag version number segment is used to identify the version information of the NEW TPID, facilitating subsequent version upgrades and compatibility. The identification steps are as follows: First, a full-bit match is performed on the NEW TPID. The received NEW TPID is compared bit by bit with the preset NEW TPID standard value. If all bits match completely, it is determined that the NEW TPID has been identified. If the full-bit match fails, the general CSIG TPID identifier segment in the NEW TPID is extracted and matched with the preset general CSIG TPID identifier standard value. If the match is successful, it is determined that the general CSIG TPID identifier has been identified. At this time, the network forwarding device processes the CSIG tag according to the processing method specified by the native CSIG protocol. If all matches fail, it is determined that no valid protocol identifier has been identified, and the telemetry message is processed according to the conventional message forwarding logic.

[0029] The third step is to read the trigger flag bit and the acquisition trigger counter TTC value in the CSIG tag.

[0030] If a NEW TPID is detected, the processor reads the trigger flag and TTC value from the CSIG tag. The trigger flag is used to identify the current telemetry trigger status, which is divided into a counted state (status value 1) and a triggered state (status value 0). The TTC value is the current value of the acquisition trigger counter, which is used to control the trigger position and timing of telemetry data acquisition. Its initial value is set by the source device according to the path hop count of the network forwarding device that needs to acquire telemetry data issued by the management node.

[0031] Fourth step: Perform the corresponding operation based on the trigger flag and TTC value: If the trigger flag is in the counting state (state value 1), the TTC value is decremented by 1 to obtain the decremented TTC value. If the decremented TTC value is 0, the trigger flag is set to the triggered state (state value 0), and the telemetry data acquisition process is initiated to collect telemetry data from the current network forwarding device. The telemetry data includes the signal value corresponding to the signal type in the CSIG tag (such as congestion signal value, bandwidth occupancy signal value, etc.) and the current location metadata (such as the device identifier and port number of the network forwarding device). After the acquisition is completed, the telemetry data is written to the signal value and location metadata fields of the CSIG tag, the CSIG tag content is updated, and then the telemetry message is forwarded to the next level device or the destination device. If the TTC value after decrementing by 1 is greater than 0, the trigger flag is kept in the counting state, the TTC value after decrementing by 1 is updated to the TTC field in the CSIG tag, and then the telemetry message is forwarded without telemetry data collection.

[0032] If the trigger flag is in the triggered state (state value is 0), no additional processing is performed on the telemetry message, and transparent forwarding is directly executed. That is, the content of the telemetry message (including the telemetry data in the CSIG tag) remains unchanged and is forwarded to the next level device or the destination device.

[0033] In this embodiment, the CSIG tag includes a compact format tag and an extended format tag, which can adopt two parallel field reuse schemes. The processor selects one of the schemes to perform the corresponding operation according to the preset global unified configuration. The specific bit definitions of the two schemes are as follows: The compact mode CSIG tag field format in the first field reuse scheme (preferred field reuse scheme) is as follows: Figure 3 As shown in the figure, the CSIG tag format in the native compact mode is compared with the CSIG tag format in the improved compact mode of this invention to more clearly distinguish them. The improved compact format CSIG tag of this invention has the same length as the native compact mode CSIG tag, with a total bit length of 32 bits, specifically defined as follows: The first 15 bits are NEW TPID, which is used to identify the new CSIG tag protocol identifier defined in this invention. NEW TPID indicates that the CSIG tag is a CSIG tag that supports the TTC hop count triggered telemetry data acquisition mechanism. Bits 16-18 are a 3-bit signal type field (signal, Type, T), used to identify the type of telemetry signal (such as congestion signal, bandwidth signal, etc.). The 19th bit is a 1-bit trigger flag (F). This field uses the R field in the original CSIG label. A status value of 1 indicates the counting state, and 0 indicates the triggered state. Bits 20-24 are a 5-bit signal value field (S), used to store the signal value corresponding to the signal type.

[0034] Bits 25-31 are 7-bit dual-state multiplexed fields used for TTC / LM multiplexing. When the trigger flag is 1, it is used as the TTC counter field to record the TTC value. When the trigger flag is 0, it is used as the location metadata (Locator Metadata, LM) field. In the first field reuse scheme (preferred field reuse scheme), the extended mode CSIG tag field format is as follows: Figure 4As shown in the figure, the CSIG tag format in the native extended mode is compared with the CSIG tag format in the improved extended mode of this invention to more clearly distinguish them. The total bit length of the extended mode CSIG tag in the first field multiplexing scheme is 64 bits, specifically defined as follows: Bits 0-15 of the first 4 bytes are NEW TPID, which indicates that the CSIG tag is a CSIG tag that supports the TTC hop count triggered telemetry data acquisition mechanism; Bits 16-31 of the first 4 bytes are a 16-bit dual-state multiplexed field used for TTC / LM multiplexing. When the trigger flag is 1, it is used as an TTC counter, and when the trigger flag is 0, it is used as the location metadata LM field. Bits 0-3 of the last 4 bytes are a 4-bit signal type field (T); Bits 4-23 of the last 4 bytes constitute a 20-bit signal value field (S); The 24th bit in the last 4 bytes is a 1-bit trigger flag (F), using the highest bit in the native extended CSIG tag R field; Bits 25-31 of the last 4 bytes are a 7-bit spare extension field, which is the remaining 7 bits in the original extended CSIG tag R field, and is used for subsequent function expansion.

[0035] The second field reuse scheme (preferred field reuse scheme) uses the compact mode CSIG tag field format as follows: Figure 5 As shown in the figure, the CSIG tag format in the native compact mode is compared with the CSIG tag format in the improved compact mode of this invention to more clearly distinguish them. In the second field multiplexing scheme, the total bit length of the compact mode tag is 32 bits, specifically defined as follows: The first 15 bits are NEW TPID, which is used to identify the new CSIG tag protocol identifier defined in this invention. NEW TPID indicates that the CSIG tag is a CSIG tag that supports the TTC hop count triggered telemetry data acquisition mechanism. Bits 16-18 are a 3-bit signal type field (T), used to identify the type of telemetry signal (such as congestion signal, bandwidth signal, etc.). The 19th bit is a 1-bit trigger flag (F). This field uses the R field in the original CSIG tag. A status value of 1 indicates the counting state and 0 indicates the triggered state. Bits 20-24 are 5-bit dual-state multiplexed fields used for TTC / S field multiplexing. When the trigger flag is 1, it is used as the TTC counter field to record the TTC value. When the trigger flag is 0, it is used to store the signal value corresponding to the signal type. The 25th to 31st positions are 7-bit location metadata LM fields.

[0036] In the second field reuse scheme, the extended mode CSIG tag field format is as follows: Figure 6 As shown in the figure, the CSIG tag format in the native extended mode is compared with the CSIG tag format in the improved extended mode of this invention to more clearly distinguish them. The total bit length of the extended mode CSIG tag in the second field multiplexing scheme is 64 bits, specifically defined as follows: Bits 0-15 of the first 4 bytes are NEW TPID, which indicates that the CSIG tag is a CSIG tag that supports the TTC hop count triggered telemetry data acquisition mechanism; Bits 16-31 of the first 4 bytes are the 16-bit location metadata LM field; Bits 0-3 of the last 4 bytes are a 4-bit signal type field (T); Bits 4-23 of the last 4 bytes are a 20-bit dual-state multiplexed field, used for TTC / S field multiplexing. When the trigger flag is 1, it is used as an TTC counter; when the trigger flag is 0, it is used to store the acquired signal value.

[0037] The 24th bit in the last 4 bytes is a 1-bit trigger flag (F), using the highest bit in the native extended CSIG tag R field; Bits 25-31 of the last 4 bytes are a 7-bit spare extension field, which is the remaining 7 bits in the original extended CSIG tag R field, and is used for subsequent function expansion.

[0038] Furthermore, telemetry messages include service messages and dedicated probe messages. When the processor executes executable instructions, it processes these two types of messages as follows: If the telemetry message is a service message (i.e., a message carrying actual service data), after the telemetry data is collected, the preset telemetry data recovery mode is determined: if the preset mode is direct reporting, the telemetry data is directly reported to the management node through the out-of-band management channel (a dedicated management channel independent of the service data transmission channel), without needing to embed the CSIG tag into the service message sent to the downstream device; if the preset mode is regular reporting, the steps of writing the telemetry data into the CSIG tag and forwarding the telemetry message are executed, and the destination device extracts the telemetry data and reports it to the management node.

[0039] If the telemetry message is a dedicated probe message (i.e., a message used only for network telemetry detection and not carrying actual business data), after the network forwarding device collects telemetry data, it constructs a CSIG feedback message, copies the collected telemetry data to the CSIG tag of the feedback message, and then sends the feedback message back to the source device or management node. Alternatively, the telemetry data can be directly reported to the management node through the out-of-band management network without forwarding the original dedicated probe message.

[0040] Example 2 like Figure 2 As shown, this embodiment provides a CSIG telemetry method based on hop count triggering, applied to the network forwarding device in Embodiment 1 above, specifically including the following steps: S1. Receive telemetry messages carrying the CSIG tag; The network forwarding device receives telemetry messages through the network interface. These telemetry messages are constructed and sent by the source device and carry a CSIG tag. The CSIG tag contains core fields such as NEW TPID, trigger flag, and TTC value.

[0041] S2. Identify the NEW TPID in the CSIG tag; NEW TPID uses a segmented format, including a general CSIG TPID identifier segment and a tag version number segment. The identification steps are as follows: S21. Perform a full-bit match on the NEW TPID. Compare the received NEW TPID with the preset NEW TPID standard value bit by bit. If the match is successful, proceed to step S3. S22. If the overall match fails, extract the general CSIG TPID identifier segment from the NEW TPID and match it with the preset general CSIG TPID identifier standard value. If the match is successful, process the CSIG label according to the native CSIG protocol; if the match fails, forward the message as usual.

[0042] S3. Read the trigger flag bit in the CSIG tag and acquire the trigger counter TTC value; The trigger flag is divided into a counting state (1) and a triggered state (0), and the TTC value is the current acquisition trigger counter value.

[0043] S4. Determine the status of the trigger flag: S41. If the trigger flag is in the counting state, decrement the TTC value by 1 to obtain the decremented TTC value; S411. Determine if the TTC value after subtracting 1 is 0; S4111. If the TTC value after decrementing by 1 is 0, the trigger flag position will be set to the triggered state, telemetry data (including the signal value corresponding to the signal type and the current location metadata) will be collected, the telemetry data will be written into the signal value and location metadata fields of the CSIG tag, and the telemetry message will be forwarded. S4112. If the TTC value after decrementing by 1 is greater than 0, maintain the trigger flag in the counting state, update the TTC value in the CSIG tag, and forward the telemetry message; S42. If the trigger flag is in the triggered state, perform transparent forwarding of telemetry messages.

[0044] In this embodiment, the two field reuse schemes of the CSIG tag and the processing method of telemetry messages (service messages and dedicated probe messages) are consistent with the network forwarding device in Embodiment 1, and will not be repeated here.

[0045] Example 3 like Figure 7 As shown, this embodiment provides a CSIG telemetry system based on hop count triggering, including a management node, source devices (source host), network forwarding devices (which may include rack-top ToR switches, aggregation Aggro switches, core switches, etc., depending on the networking environment), and destination devices (destination host). The devices establish communication connections with each other through the network to realize the collection, transmission, and reporting of telemetry data.

[0046] The management node uses a server or dedicated management equipment to implement the overall configuration and management of the telemetry system. Its specific functions are as follows: 1. Path detection: Using the Traceroute mechanism of the TCP / IP protocol, the total number of hops of the probe path from the source device to the destination device is detected, that is, the number of network forwarding devices in the path, providing a basis for setting the initial value of TTC; 2. Configuration Distribution: Send telemetry configuration commands to the source device. The commands include configuration information such as total path hops, CSIG tag field reuse scheme, and telemetry data recycling mode. 3. Dynamic adaptation: If a route change is detected through the Traceroute mechanism, the total hop count of the new path is obtained again, and the updated telemetry configuration command is sent to the source device to ensure that the initial TTC value matches the new path. 4. Data Reception and Processing: Receive telemetry data reported by destination devices or network forwarding devices, bind the telemetry data to the corresponding path nodes (network forwarding devices) based on the total number of hops, and reconstruct a complete network path view to facilitate real-time monitoring of network status and location of faulty nodes by staff.

[0047] The source device uses a network terminal or dedicated detection equipment to construct and send telemetry messages, with the following specific functions: 1. Message Construction: Based on the telemetry configuration instructions issued by the management node, construct a telemetry message carrying a CSIG tag. The CSIG tag is set with an initial TTC value that matches the total number of hops in the path (usually the initial TTC value is equal to the number of hops corresponding to the network forwarding device at the location to be probed, to ensure that the message triggers telemetry data collection at the network forwarding device at the specified location). 2. Message encapsulation and transmission: Telemetry messages are encapsulated using a fixed five-tuple (source IP address, destination IP address, source port, destination port, protocol type) to ensure stable transmission of messages in the network. The encapsulated telemetry messages are then sent to the network forwarding device.

[0048] The network forwarding device adopts the network forwarding device in Embodiment 1, and is used to execute the CSIG telemetry method based on hop count triggering in Embodiment 2, so as to realize the reception of telemetry messages, CSIG tag identification, TTC counting, telemetry data acquisition and forwarding / reporting.

[0049] The destination device receives telemetry messages and reports telemetry data. Its specific functions are as follows: 1. Message reception: Receive telemetry messages forwarded by network forwarding devices; 2. Data Extraction and Reporting: Extract telemetry data from the CSIG tag of the telemetry message, and report the telemetry data to the management node through the out-of-band management channel to complete the retrieval of telemetry data.

[0050] Furthermore, the system's processing flow for different types of telemetry messages is as follows: If the telemetry message is a service message, the network forwarding device processes it in the manner described in Example 1. If it is a regular reporting mode, the telemetry data is written to the CSIG tag and forwarded to the destination device. The destination device extracts the telemetry data and reports it to the management node. If it is a direct reporting mode, the network forwarding device directly reports the telemetry data to the management node.

[0051] If the telemetry message is a dedicated probe message, after triggering telemetry data acquisition, the network forwarding device constructs a CSIG feedback message, copies the telemetry data to the CSIG tag of the feedback message, and then sends it back to the source device or management node. Alternatively, it can be directly reported to the management node through the out-of-band management network, without needing to forward the dedicated probe message to the destination device, thus improving telemetry efficiency.

[0052] In this embodiment, the various devices of the telemetry system work together to achieve dynamic adjustment of telemetry configuration, accurate collection and efficient reporting of telemetry data, and real-time monitoring of network path status, thus solving the problems of inflexible configuration and poor adaptability of routing switching in existing telemetry systems.

[0053] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A network forwarding device, comprising: Includes a processor and a memory, the memory storing executable instructions, the processor executing the executable instructions to perform the following steps: Receive telemetry messages carrying congestion signal tags, i.e., CSIG tags, and identify the new protocol identifier, i.e., NEW TPID, in the CSIG tags; If the NEW TPID is detected, read the trigger flag bit in the CSIG tag and collect the trigger counter TTC value; If the trigger flag is in a counting state, the TTC value is decremented by 1. If the TTC value after decrementing is 0, the trigger flag is set to the triggered state, the telemetry data acquisition process is started, the telemetry data of the current network forwarding device is collected in real time, the telemetry data is written into the CSIG tag and the telemetry message is forwarded, and the CSIG tag maintains a fixed length throughout the forwarding process; if the TTC value after decrementing is greater than 0, the trigger flag is kept in a counting state and the TTC value in the CSIG tag is updated. If the trigger flag is in the triggered state, transparent forwarding is performed on the telemetry message, without modifying any content in the CSIG tag.

2. The network forwarding device according to claim 1, characterized in that, The NEW TPID adopts a segmented format, including a general CSIG TPID identifier segment and a tag version number segment; when the processor executes the executable instructions, the steps for identifying the NEW TPID specifically include: The NEW TPID is first matched with all bits. If the match is successful, the NEW TPID is identified. If the match is unsuccessful, the generic CSIG TPID identifier segment is matched. If the generic CSIG TPID identifier segment is matched successfully, the generic CSIG TPID identifier is identified. The network forwarding device processes the CSIG tag in the same way as the generic CSIG tag.

3. The network forwarding device of claim 1, wherein, The CSIG tag includes a compact format tag and an extended format tag, and the CSIG tag adopts the following first field reuse scheme: The compact format tag's bit structure is defined as follows: 16-bit NEW TPID, 3-bit signal type field, 1-bit trigger flag, 5-bit signal value field, and 7-bit dual-state multiplexing field; wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is in the "counting state" value; and serves as a location metadata LM field when the trigger flag is in the "triggered state" value. The extended format tag's bit configuration is as follows: 16-bit NEW TPID, 16-bit dual-state multiplexing field, 4-bit signal type field, 20-bit signal value field, 1-bit trigger flag, and 7-bit spare extended field; wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is in the "counting state" value, and as a location metadata LM field when the trigger flag is in the "triggered state" value.

4. The network forwarding device of claim 1, wherein, The CSIG tag includes a compact format tag and an extended format tag, and the CSIG tag adopts the following second field reuse scheme: The compact format tag's bit structure is defined as follows: 16-bit NEW TPID, 3-bit signal type field, 1-bit trigger flag, 5-bit dual-state multiplexing field, and 7-bit location metadata field; wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is in the "counting state" value, and as a signal value field when the trigger flag is in the "triggered state" value; The extended format tag's bit configuration is as follows: 16-bit NEW TPID, 16-bit location metadata field, 4-bit signal type field, 20-bit dual-state multiplexing field, 1-bit trigger flag, and 7-bit spare extended field; wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is in the "counting state" value, and as a signal value field when the trigger flag is in the "triggered state" value.

5. The network forwarding device of claim 3, wherein, When the processor executes the executable instruction, it selects the first field reuse scheme according to the preset global unified configuration and performs the corresponding operation.

6. The network forwarding device of claim 1, wherein, The telemetry messages include service messages and dedicated detection messages; If the telemetry message is a service message, then after telemetry data collection, it also includes: Determine the preset telemetry data collection mode. If it is the direct reporting mode, the telemetry data is directly reported to the management node through the out-of-band management channel. If it is the normal reporting mode, the steps of writing the telemetry data into the CSIG tag and forwarding the telemetry message are executed so that the destination device can report. If the telemetry message is a dedicated probe message, after the network forwarding device collects telemetry data, it constructs a CSIG feedback message, copies the telemetry data to the CSIG tag of the feedback message, and then sends it back to the source device or management node, or directly reports it to the management node through the out-of-band management network.

7. A CSIG telemetry method based on hop count triggering, applied to a network forwarding device, characterized in that, Includes the following steps: Receive telemetry messages carrying congestion signal tags, i.e., CSIG tags, and identify the new protocol identifier, i.e., NEW TPID, in the CSIG tags; If the NEW TPID is detected, read the trigger flag bit in the CSIG tag and collect the trigger counter TTC value; If the trigger flag is in a counting state, the TTC value is decremented by 1. If the TTC value after decrementing is 0, the trigger flag is set to the triggered state, the telemetry data acquisition process is started, the telemetry data of the current network forwarding device is collected in real time, the telemetry data is written into the CSIG tag and the telemetry message is forwarded, and the CSIG tag maintains a fixed length throughout the forwarding process; if the TTC value after decrementing is greater than 0, the trigger flag is kept in a counting state and the TTC value in the CSIG tag is updated. If the trigger flag is in the triggered state, transparent forwarding is performed on the telemetry message, without modifying any content in the CSIG tag.

8. The method according to claim 7, characterized in that, The NEW TPID adopts a segmented format, including a general CSIG TPID identifier segment and a tag version number segment; the specific steps for the network forwarding device to identify the NEW TPID include: The NEW TPID is first matched with all bits. If the match is successful, the NEW TPID is identified. If the match is unsuccessful, the generic CSIG TPID identifier segment is matched. If the generic CSIG TPID identifier segment is matched successfully, the generic CSIG TPID identifier is identified. The network forwarding device processes the CSIG tag in the same way as the generic CSIG tag.

9. The method of claim 7, wherein, The CSIG tag includes a compact format tag and an extended format tag, and the CSIG tag adopts the following first field reuse scheme: The compact format tag's bit structure is defined as follows: 16-bit NEW TPID, 3-bit signal type field, 1-bit trigger flag, 5-bit signal value field, and 7-bit dual-state multiplexing field; wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is 1, and as a location metadata LM field when the trigger flag is 0. The extended format tag's bit structure is defined as follows: 16-bit NEW TPID, 16-bit dual-state multiplexing field, 4-bit signal type field, 20-bit signal value field, 1-bit trigger flag, and 7-bit spare extended field; wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is 1, and as a location metadata LM field when the trigger flag is 0.

10. The method of claim 7, wherein, The CSIG tag includes a compact format tag and an extended format tag, and the CSIG tag adopts the following second field reuse scheme: The compact format tag's bit structure is defined as follows: 16-bit NEW TPID, 3-bit signal type field, 1-bit trigger flag, 5-bit dual-state multiplexing field, and 7-bit location metadata field; wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is 1, and as a signal value field when the trigger flag is 0. The extended format tag's bit configuration is as follows: 16-bit NEW TPID, 16-bit location metadata field, 4-bit signal type field, 20-bit dual-state multiplexing field, 1-bit trigger flag, and 7-bit spare extended field; wherein, the dual-state multiplexing field serves as a TTC counter when the trigger flag is 1, and as a signal value field when the trigger flag is 0.

11. The method according to claim 8, characterized in that, The network forwarding device selects the first field reuse scheme and performs the corresponding operation according to the preset global unified configuration.

12. The method of claim 7, wherein, The telemetry messages include service messages and dedicated detection messages; If the telemetry message is a service message, then after telemetry data collection, it also includes: Determine the preset telemetry data collection mode. If it is the direct reporting mode, the telemetry data is directly reported to the management node through the out-of-band management channel. If it is the normal reporting mode, the steps of writing the telemetry data into the CSIG tag and forwarding the telemetry message are executed so that the destination device can report. If the telemetry message is a dedicated probe message, after the network forwarding device collects telemetry data, it further includes: constructing a CSIG feedback message, copying the telemetry data to the CSIG tag of the feedback message and sending it back to the source device or management node, or directly reporting it to the management node through the out-of-band management network.

13. A CSIG telemetry system based on hop count triggering, the system comprising: This includes management nodes, source devices, network forwarding devices, and destination devices; The management node is used to detect the total number of hops of the probe path from the source device to the destination device through the Traceroute mechanism of the TCP / IP protocol, and send telemetry configuration commands to the source device. The source device is configured to construct a telemetry message carrying a CSIG tag according to the telemetry configuration instruction, set the initial value of TTC in the CSIG tag, encapsulate it with a fixed 5-tuple, and send it to the network forwarding device. The CSIG tag is a fixed-length tag. The network forwarding device is used to execute the CSIG telemetry method based on hop count triggering as described in any one of claims 7-12; The destination device is used to receive the telemetry message, extract the telemetry data from the CSIG tag, and report it to the management node through the out-of-band management channel.

14. The system of claim 13, wherein, The telemetry messages include service messages and dedicated detection messages; If the telemetry message is a service message, the destination device extracts telemetry data from the telemetry message and reports it to the management node through the out-of-band management channel; If the telemetry message is a dedicated probe message, after triggering the collection, the network forwarding device constructs a CSIG feedback message, copies the telemetry data to the CSIG tag of the feedback message, and then sends it back to the source device or management node, or directly reports it to the management node through the out-of-band management network.

15. The system of claim 14, wherein, The management node is also used for: If a route change is detected due to a route switch, the new total path hop count is obtained again through the Traceroute mechanism, and an updated telemetry configuration command is sent to the source device. Receive telemetry data reported by the destination device or network forwarding device, bind the telemetry data to path nodes based on the total number of hops on the path, and reconstruct a complete path view.

Images