Message pruning method based on survival time field, electronic device and storage medium

Abstract

The application discloses a message pruning method based on a survival time field, an electronic device and a storage medium, and belongs to the technical field of communication. The method comprises the following steps: in response to the occurrence of congestion packet loss, pruning a first message received based on a survival time field to obtain a second message; modifying the value of the survival time field of the second message to a first value, wherein the first value is used to indicate that the second message is a pruned message; and storing the second message in a transmission queue for transmission.

Description

Technical Field

[0001] This application relates to the field of communication technology, and in particular to a message trimming method, electronic device, and storage medium based on a time-to-live field. Background Technology

[0002] In related technologies, Differentiated Services Code Points (DSCPs) are used to identify trimmed and trimmed packets. However, in a wide area network (WAN) environment, using DSCPs to identify trimmed and trimmed packets may lead to the following problems: 1) Since DSCP values ​​are used in wide area networks to configure Quality of Service (QoS) policies, a portion of DSCP values ​​are reserved to identify potential conflicts between clipped packets and existing network QoS policies. 2) The number of DSCP priority fields is limited. For example, for intelligent computing data center networks, network traffic can be planned, and a portion of the DSCP field can be proactively allocated for packet pruning. However, for wide area networks, managing the allocation of DSCP fields is very complex. 3) Reserved DSCP values ​​for pruning require unified configuration of network devices. If some devices do not support the specified DSCP values, the pruned packets may not be correctly identified and processed (e.g., they may be dropped).

[0003] Therefore, how to trim packets in a wide area network is a technical problem that needs to be solved in related technologies. Summary of the Invention

[0004] This application provides a message trimming method, electronic device, and storage medium based on a time-to-live field, which can at least solve one of the aforementioned problems caused by using DSCP to identify trimmable messages and trimmed messages.

[0005] To solve the above-mentioned technical problems, this application is implemented as follows: In a first aspect, a message trimming method based on a time-to-live (TTL) field is provided, applied to a network node. The method includes: in response to congestion and packet loss, trimming a received first message based on the TTL field to obtain a second message; modifying the value of the TTL field of the second message to a first value, wherein the first value is used to indicate that the second message is a trimmed message; and storing the second message in a transmission queue for transmission.

[0006] Secondly, a message identification method is provided, applied to a receiving end. The method includes: receiving a target message; obtaining the value of the time-to-live (TTL) field of the target message; and identifying the target message as a truncated message in response to the TTL field indicating that the target message is a truncated message.

[0007] Thirdly, a message transmission method is provided, the method comprising: acquiring a target message to be entered into a network; in response to the target message being a message that can be pruned, setting the value of the time-to-live field of the target message to a second value, wherein the second value is used to indicate that the target message is a message that can be pruned; and transmitting the target message to a network node.

[0008] Fourthly, an electronic device is provided, the electronic device including a processor and a memory, the memory storing a program or instructions executable on the processor, the program or instructions, when executed by the processor, implementing the steps of the method described in the first, second, or third aspects above.

[0009] Fifthly, a readable storage medium is provided, on which a program or instructions are stored, which, when executed by a processor, implement the steps of the method described in the first, second, or third aspects above.

[0010] In a sixth aspect, a computer program product is provided, the computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program comprising program instructions that, when executed by a computer, cause the computer to perform the steps of the method described in the first, second, or third aspect above.

[0011] In this embodiment, in response to congestion and packet loss, the received first packet can be trimmed based on the time-to-live (TTL) field to obtain a second packet; the value of the TTL field of the second packet is modified to a first value indicating that the second packet is a trimmed packet; finally, the second packet is stored in the transmission queue for transmission. Using the TTL field to trim received packets not only reduces the impact on existing network QoS policies but also reduces network management complexity.

[0012] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0013] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0014] Figure 1 A schematic diagram of the package cutting technology in the related art is shown; Figure 2 The illustration shows a flowchart of a message trimming method based on a time-to-live field provided in an exemplary embodiment of this application; Figure 3 This illustration shows a schematic diagram of an identification message that has been truncated according to an exemplary embodiment of this application; Figure 4 The illustration shows a flowchart of another message trimming method based on a time-to-live field provided in an exemplary embodiment of this application; Figure 5 A flowchart illustrating a message identification method provided in an exemplary embodiment of this application is shown. Figure 6 A flowchart illustrating a message transmission method provided in an exemplary embodiment of this application is shown. Figure 7 This illustration shows a schematic diagram of setting up TTLTrimmable based on an exemplary embodiment of this application; Figure 8 This illustration shows a schematic diagram of a packet pruning function implemented in a network based on TTL, provided by an exemplary embodiment of this application. Figure 9 This illustration shows a schematic diagram of a packet trimming function processing mechanism based on control configuration provided in an exemplary embodiment of this application; Figure 10 A structural block diagram of an electronic device provided in an exemplary embodiment of this application is shown. Detailed Implementation

[0015] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0016] Network device port bandwidth doubles every 18 months and is likely to continue increasing rapidly in the coming years. It is projected that by 2030, port bandwidth in intelligent computing center networks will primarily be 1.6T and 3.2T. However, the on-chip cache space of switches has not increased accordingly. Furthermore, some application models (such as synchronous burst aggregation during large model training) may lead to cache resource exhaustion, resulting in congestion and packet loss. Currently, switches operate in two modes: (1) Lossless mode: The lossless mode based on priority flow control (PFC) reduces packet loss, but may bring other problems (PFC storm, deadlock).

[0017] (2) Lossy Mode (best effort): For the traditional Transmission Control Protocol (TCP), packet loss or out-of-order delivery is determined based on the Out-of-Order (OOO) threshold (e.g., receiving three identical Acknowledgement (ACK) characters). With this fast recovery design, performance is generally good as long as the tail packet is not lost. However, for tail packet loss, a timeout mechanism is required. For the Ultra Ethernet Transport (UET) layer, packet spraying technology is used to improve load balancing. In this case, out-of-order delivery is more prominent. To avoid frequent packet retransmissions, the OOO threshold needs to be increased. In this situation, the detection and response speed for packet loss and retransmission caused by congestion is correspondingly slower.

[0018] UET improves packet loss detection speed by using packet trimming technology, such as... Figure 1 As shown, packet trimming technology, when congestion and packet loss occur, instead of simply discarding the packet, it trims the packet while retaining the header and necessary payload (e.g., 128 bytes, the specific size must be considered based on the UET header recommendation) and sends it to the receiver with higher priority. Because the trimmed packet has a higher priority, the packet loss signal can reach the receiver faster, and the receiver quickly returns an ACK to the sender, thus speeding up the congestion control loop. Ideally, the packet loss detection time can be reduced to 1x the base round-trip time (RTT).

[0019] Supporting packet trimming requires corresponding functional upgrades to both the switch and the receiver. The switch needs to identify which packets can be trimmed (Trimmable), and the receiver needs to identify which packets are trimmed (Trimmed). In the UET standard, packet types are defined by using a special DSCP tag. Specifically, DSCP_Trimmable (e.g., priority 1) indicates that the packet can be trimmed by the switch. After trimming, the packet is marked with DSCP_Trimmed (e.g., priority 6) to indicate that it is a trimmed packet. When the receiver receives a packet marked with DSCP_Trimmed, it does not treat it as a normal service but precisely identifies it as a trimmed packet. The allocation of DSCP fields is planned by the control plane, and there is a mapping relationship between Trimmable and Trimmed DSCPs, for example, {1->6}.

[0020] When the application of packet trimming technology is expanded from data centers to wide area networks (WANs), it can also accelerate packet loss signal feedback and achieve performance gains. However, using DSCP to identify trimmed packets and trimmed packets on WANs has the following drawbacks: (1) In wide area networks, DSCP values ​​are used to configure service quality (QoS) policies, and a portion of DSCP values ​​are reserved to identify whether the clipped packets may conflict with the existing network QoS policies.

[0021] (2) The number of DSCP priority fields is limited. For intelligent computing data center networks, network traffic can be planned, and a portion of the DSCP fields can be actively allocated specifically for packet clipping. However, for wide area networks, managing the allocation of DSCP fields is very complex. (3) Reserved DSCP values ​​for pruning require unified configuration of network devices. If some devices do not support the specified DSCP values, the pruned packets may not be correctly identified and processed (e.g., they may be dropped).

[0022] In summary, DSCP in WANs already involves complex QoS policies and traffic engineering. In a WAN environment, redistributing DSCP could alter many existing network functionalities, making it less than ideal. Furthermore, since the IETF IPv4 standard is now finalized, extending or redefining existing fields may cause compatibility issues on network devices. Therefore, when considering identification schemes for WAN packet shaving, the redundancy of existing fields should be utilized as much as possible to implement packet shaving without altering the network device's processing flow and functional divisions.

[0023] To address at least one of the problems in the aforementioned related technologies, this application provides a message trimming technology solution based on a time-to-live field. The technical solution provided by this application embodiment will be described below with reference to the accompanying drawings.

[0024] Please see Figure 2 , Figure 2 This diagram illustrates a message pruning method based on a time-to-live (TTL) field, as provided in an embodiment of this application. The method can be executed by a network node. Figure 2 As shown, the method 200 may include the following steps.

[0025] Step 201: In response to congestion and packet loss, the received first packet is trimmed based on the time-to-live field to obtain the second packet.

[0026] In this embodiment of the application, in response to congestion and packet loss, the received first message can be trimmed based on the Time To Live (TTL) field to obtain the second message.

[0027] In practical applications, in the network, messages are encapsulated in the payload of Internet Protocol (IP) and forwarded. Various transport layer protocols are used for different scenarios, such as Virtual Extensible LAN (VXLAN), TCP / User Datagram Protocol (UDP), UET, and RDMA over Converged Ethernet version 2 (RoCEv2), as shown in Table 1.

[0028] Table 1

[0029] In the IP header, the TTL field is an 8-bit field with 256 possible values ​​(0-255). It represents the maximum number of hops a data packet can traverse. The TTL field is decremented by 1 with each hop, and the packet is discarded when the TTL value reaches 0. However, the maximum number of hops in a wide area network (WAN) is generally no more than 20. The default value is typically 128 in Windows systems and 64 in Linux systems. Therefore, the TTL field supports 256 values, providing redundancy.

[0030] In practical applications, when congestion and packet loss occur, gateway nodes can determine whether to prune received packets based on control plane configuration, or manually. For example, gateway nodes can verify the TTL value of packets according to control plane configuration to ensure that the packet is within or outside the TTL range of the packet pruning function.

[0031] In this embodiment, the first packet is a packet that meets the trimming requirements. In response to congestion and packet loss, the network node can trim the received first packet based on the TTL to obtain the second packet. Since the TTL is in the IP header field and has a wide range of values ​​with many redundant values, the impact on the existing network QoS policy can be reduced, and the complexity of network management can also be reduced.

[0032] Step 202: Modify the value of the Time to Live field of the second message to the first value.

[0033] The first value is used to indicate that the second message is a clipped message.

[0034] In this embodiment of the application, the value of the time-to-live field of the second message can be modified to a first value (TTL_trimmed) used to indicate that the second message is a trimmed message.

[0035] like Figure 3 As shown, when a network node trims the first packet to obtain the second packet, it can mark it with the TTL_trimmed flag to indicate that the second packet has been trimmed. This makes it easier for the receiving end to determine that the second packet has been trimmed and then perform the appropriate processing, saving time.

[0036] Step 203: Store the second message in the transmission queue for transmission.

[0037] In this embodiment of the application, the second message, after being trimmed and having its time-to-live field value modified, can be stored in a transmission queue for transmission.

[0038] In this embodiment, in response to congestion and packet loss, the received first packet can be trimmed based on the time-to-live (TTL) field to obtain a second packet; the value of the TTL field of the second packet is modified to a first value indicating that the second packet is a trimmed packet; finally, the second packet is stored in the transmission queue for transmission. This not only reduces the impact on existing network QoS policies but also reduces network management complexity.

[0039] In some embodiments, the TTL (Time to Live) can be used to indicate whether the first message can be truncated. In these embodiments, truncating the received first message based on the Time to Live field to obtain the second message includes: In response to the first message's time-to-live field taking a second value, the first message is trimmed to obtain a second message, wherein the second value is used to indicate that the first message is a message that can be trimmed.

[0040] In the above embodiments, a second value (TTL_trimmable) can be used to identify whether the first packet is a packet that can be trimmed. When congestion and packet loss occur in the network device, if the time-to-live field of the first packet is the second value TTL_trimmable, the first packet can be trimmed to obtain the second packet.

[0041] In practical applications, after a message is trimmed (assuming a TTL=255 is added), the TTL value can decrease hop by hop during transmission. If the TTL value exceeds the range of the trimmed message and the message is not discarded, it may be trimmed again when congestion is encountered again, and the TTL value of the message may be reset to the TTL_trimmed value. In this case, if a loop exists in the network, the above message may exist in the network for a long time, resulting in a waste of network resources.

[0042] In some embodiments, to prevent already trimmed packets from being trimmed again, the trimmed packets and trimmable packets may be limited to not sharing the same TTL range. In these embodiments, the first value is within a first value range, the second value is within a second value range, the first value range and the second value range do not overlap, the time-to-live field value within the first value range is used to identify the trimmed packet, and the time-to-live field value within the second value range is used to identify the packet that can be trimmed.

[0043] For example, the first value range can be set to TTL=201~255, indicating that the packet has been clipped, and the second value range can be set to TTL=170~200, indicating that the packet can be clipped. In this way, the first and second value ranges do not overlap, which can effectively prevent clipped packets from being clipped again and avoid wasting resources.

[0044] In some embodiments, the first value can be the upper bound of the first value range. In practical applications, the upper bound of TTL_Trimmed can be used to indicate that the packet is a trimmed packet.

[0045] In some embodiments, when a clipped message is received, if the TTL value of the message is greater than the lower bound of a first value range, the TTL value can be decremented by 1 before transmission. In these embodiments, the method further includes the following steps: Step 1: Received the third message; Step 2: Obtain the value of the Time to Live field of the third message; Step 3: In response to the fact that the value of the time-to-live field of the third message is greater than a preset value, the value of the time-to-live field of the third message is reduced by 1 before transmission, wherein the preset value is greater than or equal to the lower bound of the first value range and less than or equal to the upper bound of the first value range.

[0046] In the above embodiments, a third message can be received and the value of its Time-to-Live (TTL) field can be obtained. If the TTL value of the third message is greater than a preset value, it can be determined that the third message is a trimmed message. In this case, the TTL value of the third message can be decremented by 1 before transmission. For example, when a message is trimmed, a judgment logic can be added to the network node. If the current TTL value of the message is greater than a preset value (for example, it can be the lower bound of TTL_Trimmed, without restriction), the network node decrements the TTL by 1 and then transmits the message. This ensures the normal transmission of the third message.

[0047] In some embodiments, to avoid wasting resources, after obtaining the value of the time-to-live field of the third message, the method may further include: in response to the value of the time-to-live field of the third message being equal to the preset value, discarding the third message.

[0048] In the above embodiments, if the value of the time-to-live field of the third message is equal to the preset value, it indicates that the IP address of the third message is unreachable, and the third message can be discarded. This can avoid wasting resources and speed up message transmission.

[0049] In some embodiments, to alleviate network congestion, packets with a TTL value less than TTL_Trimmable may be dropped in the event of congestion-induced packet loss. In these embodiments, the method further includes the following steps: Step 1: The fourth message was received; Step 2: In response to the occurrence of congestion and packet loss, and the value of the time-to-live field of the fourth packet being less than the lower bound of the second value range, the fourth packet is discarded.

[0050] In the above embodiment, if a fourth packet with a TTL value less than the lower bound of the second value range is received (indicating that the fourth packet cannot be truncated) and congestion and packet loss occur, the fourth packet can be discarded. In this way, by discarding packets that cannot be truncated, network traffic can be adjusted to relieve network overload and help alleviate congestion problems.

[0051] In some embodiments, to ensure that the trimmed second message can be transmitted to the receiving end quickly with high priority, the DSCP of the second message can be modified to high priority. In these embodiments, storing the second message in the transmission queue for transmission may include: Step 2031: Modify the distinguishing service code point of the second message from the first priority value to the second priority value, wherein the priority corresponding to the second priority value is higher than the priority corresponding to the first priority value; Step 2032: Store the second message in the transmission queue corresponding to the priority value of the second priority value for transmission.

[0052] In the above embodiment, the DSCP of the second message can be modified from a first priority value to a second priority value, which is higher than the priority corresponding to the first priority value. Then, the second message is stored in a transmission queue with a priority corresponding to the second priority value and transmitted. This ensures timely congestion signal feedback.

[0053] To support packet trimming, network devices need to be upgraded to support TTL-based processing. Figure 4 The illustration shows a flowchart of another message pruning method based on a time-to-live field, as shown in an exemplary embodiment of this application. Figure 4 As shown, the method includes the following steps: Step 401: Start, message received.

[0054] Step 402: Is it a protocol message type? If it is determined to be a protocol message, for example, if the message type is a well-known protocol message type (such as ICMP message), you can proceed to step 4032 to follow the relevant process. If it is determined that the message is not a protocol message, you can proceed to step 4031.

[0055] Step 4031: Decrease the TTL value by 1.

[0056] Step 4032: Process according to relevant procedures. These relevant procedures can be processes involving other messages; this embodiment does not impose specific limitations.

[0057] Step 404: Is the current TTL value the specified value or 0? It can be determined whether the current TTL is a specified value (such as the lower bound of TTL_Trimmed or 0). If it is not, proceed to step 4051. If it is, it indicates that the IP of the packet is unreachable, and step 4052 can be performed to discard the current packet.

[0058] Step 4051: Has congestion and packet loss occurred? In practical applications, it can be determined whether congestion and packet loss have occurred in the network device. If so, proceed to step 406; otherwise, proceed to step 4032 and process the packets according to the congestion handling methods in the relevant process.

[0059] Step 4052: Discard the message.

[0060] Step 406: Is the current TTL Trimmable? Determine whether the current packet carries TTL_Trimmable. If it does, proceed to step 407; otherwise, proceed to step 4052 and discard the packet. This saves resources. Step 407: Prune the packet, modify the TTL to the upper bound of the Trimmed range, modify the DSCP, and store the pruned packet in a high-priority queue. Pruning the packet, modifying the TTL to the upper bound of TTL_Trimmed, modifying the DSCP to high priority, and storing the pruned packet in the corresponding high-priority queue can ensure timely congestion signal feedback.

[0061] Step 408: End.

[0062] It should be noted that the TTL-based message trimming method is suitable not only for wide area network environments, but also for local area network environments or data center environments, and the embodiments of this application do not impose any limitations on this.

[0063] It should be noted that although this application mainly describes the packet trimming method based on TTL, other similar schemes, such as those based on existing IPv4 / v6, MPSL, MAC header fields, or those based on existing IPv6 / SRv6, IPv4, etc., that carry packet trimming information by expanding the fields, are also within the scope of this application. Their implementation process is similar to the above-mentioned TTL-based packet trimming process, and will not be described in detail here.

[0064] Figure 5 A flowchart illustrating a message identification method according to an exemplary embodiment of this application is shown. This method can be executed by a receiving end. The receiving end can be an end-side host or a receiving end gateway device; this embodiment does not specifically limit the scope. Figure 5 As shown, the message identification method mainly includes the following steps.

[0065] Step 501: Target message received.

[0066] In the embodiments of this application, the receiving end can receive the target message, which can be a trimmed message or an untrimmed message.

[0067] Step 502: Obtain the value of the Time to Live field of the target message.

[0068] In practical applications, network nodes can modify the value of the Time-to-Live (TTL) field of a target packet. For example, they can modify the TTL field value to indicate that the target packet is a truncated packet. In this embodiment, the receiving end can obtain the value of the TTL field of the target packet to determine the type of the target packet and take appropriate action.

[0069] Step 503: In response to the value of the time-to-live field of the target message indicating that the target message is a truncated message, identify the target message as a truncated message.

[0070] In this embodiment, the receiving end can determine whether a target packet is a truncated packet based on the value of its time-to-live (TTL) field. If the TTL field value indicates that the target packet is a truncated packet, then the target packet is identified as such. This allows for further processing based on the truncated nature of the target packet, ensuring normal data transmission and avoiding resource waste.

[0071] In this embodiment, the receiving end can receive the target packet and then obtain the value of the time-to-live (TTL) field of the target packet. Finally, in response to the TTL field value indicating that the target packet is a truncated packet, the receiving end identifies the target packet as a truncated packet. In this way, the receiving end can determine whether the received target packet is a truncated packet and perform corresponding processing, which helps ensure normal data transmission and avoids wasting resources.

[0072] In some embodiments, the target message can be processed according to the performance of the receiving end. For example, a receiving end with packet trimming capabilities can perform corresponding packet trimming on the target message, while a receiving end without such capabilities can discard the received target message. In these embodiments, after identifying the target message as a trimmed message, the method further includes the following steps: Step 5041: In response to having the capability to process the trimmed packets, perform packet trimming feedback processing on the target packet; or, Step 5042: In response to the lack of ability to process the truncated message, the target message is discarded.

[0073] In the above embodiments, if the receiving end has the ability to process the clipped messages, it can perform the corresponding packet rigging process, such as constructing a corresponding negative acknowledgment (NACK) message feedback, which is the same as the receiving end's processing process for clipped messages in related technologies. This application embodiment does not make specific limitations.

[0074] If the receiving end lacks the capability to process the trimmed packets, other simplification methods can be implemented, such as discarding them. In this case, the receiving end may not have the processing capability, and a proxy device can identify and process the trimmed target packets (e.g., discard them). This application embodiment does not limit the specific processing method. This can optimize overall transmission performance and avoid wasting resources.

[0075] Figure 6 A flowchart illustrating a message transmission method according to an exemplary embodiment of this application is shown. This method can be executed by a sender or an incoming gateway device for the message. Figure 6 As shown, the message transmission method mainly includes the following steps.

[0076] Step 601: Obtain the target message to be entered into the network.

[0077] In practical applications, in order to prevent the sending end from sending malicious service traffic that simulates the TTL value of the truncated packets, the gateway device can adjust the TTL value of the packets entering the network to ensure that only service packets that meet the requirements enter the network. This application does not restrict the specific boundary node or gateway control method.

[0078] In this embodiment of the application, the target message to be entered into the network is obtained.

[0079] Step 602: In response to the target message being a message that can be trimmed, set the value of the time-to-live field of the target message to a second value.

[0080] The second value is used to indicate that the target message is a message that can be pruned.

[0081] In this embodiment of the application, when the target message is a message that can be pruned, the value of the time-to-live field of the target message can be set to a second value to indicate that the target message is a message that can be pruned.

[0082] In practical applications, some devices, such as gateway nodes, can indicate which flow IDs' packets can be trimmed based on control plane configuration and mark them with the TTL_Trimmable flag. For packets that do not need to be trimmed, the gateway node can verify the packet's TTL value according to the control plane configuration to ensure that it is not within the TTL range of the packet trimming function.

[0083] Furthermore, in practical applications, customer data centers and intelligent computing networks can identify whether a packet is trimmable using the existing DSCP-based method. In wide area networks, packet trimming is implemented based on TTL. In this case, the intelligent computing gateway node can map the DSCP_Trimmable and TTL values ​​on the endpoint according to a static policy. Based on the DSCP_Trimmable value corresponding to the target packet, the time-to-live (TTL) field of the target packet is set to a second value. Various specific implementation methods exist, and this application does not impose specific limitations. This makes it easier to identify target packets as trimmable, helping to solve congestion problems.

[0084] Step 603: Transmit the target message to the network node.

[0085] In this embodiment of the application, the target message can be transmitted to the network node to ensure the normal transmission of the target message.

[0086] In this embodiment, a target packet to be entered into the network can be obtained. Then, in response to the target packet being a pruningable packet, the time-to-live (TTL) field of the target packet is set to a second value indicating that the target packet is a pruningable packet. Finally, the target packet is transmitted to the network node. This allows the target packet to be identified as a pruningable packet, which helps to solve congestion problems and ensures the normal transmission of the target packet.

[0087] In some embodiments, the range of the second value can be limited to distinguish between clipped packets and packets that can be clipped. In these embodiments, the second value is within a second value range, which does not overlap with a first value range. The time-to-live (TTL) field value within the first value range is used to identify clipped packets, and the TTL field value within the second value range is used to identify packets that can be clipped.

[0088] In practical applications, the first value range can be TTL=201~255, and the second value range can be TTL=170~200. In this way, since the second value range does not overlap with the first value range, that is, the clipped message and the message that can be clipped do not share the TTL range, it can prevent the clipped message from being clipped again during subsequent transmission, and also avoid wasting resources.

[0089] The technical solutions provided in the embodiments of this application are described below through specific examples.

[0090] Example 1 In this embodiment, TTL Trimmable can be configured on the intelligent computing network gateway device. Using this method, the customer data center and the intelligent computing network can identify whether a packet is trimmable using the existing DSCP-based method, while packet trimming is implemented on the wide area network based on TTL. The intelligent computing gateway device has the ability to convert between DSCP_Trimmable and TTL_Trimmable, as well as between DSCP_Trimmed and TTL_Trimmed.

[0091] like Figure 7 As shown, assuming DSCP_trimmable=1 in the customer's DC network and DSCP_Trimmed=6 in the intelligent computing center network, when the intelligent computing CPE gateway receives a message with DSCP=1, it can add TTL_Trimmable information to it. Assuming the range of TTL_trimmable is 170~200, the intelligent computing CPE gateway can set the message's TTL to 200. For messages without DSCP=1, the message's TTL is checked to ensure that the message's TTL value is not within the range of TTL_Trimmable (170~200).

[0092] In the above embodiments, when a trimmable message (TTL_Trimmable) is transmitted in the network, the following situations may occur: (1) If congestion occurs in the WAN and packet loss occurs, and the TTL value of the packet is in the range of 170~200, the current packet is trimmed and the TTL of the packet is modified to the upper limit of TTL_Trimmed. For example, if the TTL_Trimmed range of the packet is 220~250, the TTL value can be set to 250. The trimmed packet DSCP can be set to a higher priority, for example, DSCP=56. The setting of DSCP priority is entirely based on the QoS policy of the WAN and is unrelated to the division of DSCP by the data center network on the end side.

[0093] (2) If the message does not experience congestion or packet loss in the wide area network, the TTL is decremented by 1 for each hop. If no loop is encountered, the message arrives at the receiving end normally and is transmitted correctly. (3) If the message does not experience congestion and packet loss in the wide area network, but encounters a loop, the TTL is reduced by 1 for each hop. When the TTL is reduced to 0, the message is discarded. In this case, the message is discarded for reasons other than congestion. The sender and receiver can retransmit the message based on the traditional packet loss judgment logic.

[0094] In the above embodiments, the trimmed message (TTL_trimmed) is transmitted in the network in the following ways: (1) If there is no reachable route or a routing loop exists, the TTL of the trimmed packet will be decremented by 1 for each hop. When the network device determines that the TTL value is 220 (the lower limit of TT_trimmed), it can actively discard the trimmed packet. At this time, the transport layer relies on the original timeout or NACK confirmation mechanism to retransmit the packet.

[0095] (2) If the clipped message arrives at the receiving end normally, when the outgoing intelligent computing gateway receives a message with the TTL_Trimmed tag (e.g., the current TTL=245), it will identify the message as a clipped message, set the message's DSCP to 6, and the receiving end will identify the clipped message based on DSCP=6 and perform corresponding processing.

[0096] Example 2 This embodiment illustrates a method for implementing packet trimming based on TTL in a more general network. Assume there are two traffic streams in the network: host1 sends a stream to host3, and host2 sends a stream to host4. Host4 has the ability to identify and process trimmed packets, and host3 also needs to have the ability to identify trimmed packets (simply by judging based on the TTL value range), but may lack the processing capability due to resource limitations, protocol issues, or other factors.

[0097] like Figure 8 As shown, the gateway device cannot distinguish which flow is trimmable traffic and which is non-trimmable traffic; this is a capability supported by the end-side transport layer. In the gateway device, only simple policy configuration is performed: ensuring that the TTL value of service packets (without processing ICMP and other protocol packets) is less than the TTL_Trimmed range, assuming it's 220-250. If the TTL value of a packet exceeds 220, it is adjusted to a value below 220. For network nodes, since it cannot distinguish whether incoming network traffic is trimmable, when congestion and packet loss occur, all packets with a TTL less than the TTL_Trimmed range are trimmed. Setting TTL=250 and dscp=56 means that both flow1 and flow2 are trimmed. The outgoing gateway does not perform any special processing on packets and can forward them to the receiving end. When the receiving end host4 identifies a packet as trimmed, it performs corresponding processing such as packet trimming feedback. For host3, when the TTL value is determined to be Trimmed, it is discarded.

[0098] Example 3 In this embodiment, an example is given to illustrate the packet trimming function processing mechanism based on management configuration.

[0099] like Figure 9As shown, the basic assumptions are similar to those in Example 2, namely, flow1 is normal traffic and flow2 is trimmable traffic. This information can be identified by configuring the gateway device through the management plane. For flow2 packets, the TTL is set to TTL_Trimmable (e.g., 200), and for flow1 packets, the TTL value must be less than the TTL_Trimmable range (e.g., less than 170).

[0100] When network congestion occurs, packets from flow1 are dropped. For packets from flow2, assuming they carry a TTL_Trimmable of 190, they are trimmed and their TTL is set to 250, with dscp=56. If, when congestion occurs, the TTL of packets from flow2 has already dropped below 170, it indicates that the network is unreachable. In this case, no further special packet trimming is needed; the packet is simply dropped, relying on the transport layer timeout retransmission mechanism.

[0101] No special processing is required on the outgoing gateway device. Since the flow1 packet is not truncated, the host3 receiving end does not need to modify the TTL range for identification. When the receiving end Host4 identifies the packet as a truncated packet, it performs corresponding processing such as packet truncating feedback, similar to Example 2, and will not be described again.

[0102] like Figure 10 As shown, this application embodiment also provides an electronic device 1000, including a processor 1001 and a memory 1002. The memory 1002 stores a program or instructions that can run on the processor 301. When the program or instructions are executed by the processor 1001, they implement the various steps of the above-described uplink control information resource configuration method embodiment and can achieve the same technical effect. To avoid repetition, they will not be described again here.

[0103] In one exemplary embodiment, a readable storage medium is also provided, which stores a program or instructions. The stored program or instructions, via a processor, executes all or part of the steps in the message trimming method implementing the aforementioned time-to-live field, or all or part of the steps in the aforementioned message identification method, or all or part of the steps in the aforementioned message transmission method. For example, the readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, floppy disk, or optical data storage device, etc.

[0104] In one exemplary embodiment, a computer program product is also provided, the computer program product including a computer program stored on a non-transitory readable storage medium, the computer program including program instructions, which, when executed by a computer, cause the computer to perform all or part of the steps in the above-described message trimming method for the time-to-live field, or to implement all or part of the steps in the above-described message identification method, or to implement all or part of the steps in the above-described message transmission method.

[0105] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the claims.

[0106] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A message trimming method based on a time-to-live (TTL) field, characterized in that, Applied to network nodes, including: In response to congestion and packet loss, the received first packet is trimmed based on the time-to-live field to obtain the second packet; Modify the value of the time-to-live field of the second message to a first value, wherein the first value is used to indicate that the second message is a clipped message; The second message is stored in the transmission queue for transmission.

2. The method according to claim 1, characterized in that, The process of trimming the received first message based on the time-to-live field to obtain the second message includes: In response to the first message's time-to-live field taking a second value, the first message is trimmed to obtain a second message, wherein the second value is used to indicate that the first message is a message that can be trimmed.

3. The method according to claim 2, characterized in that, The first value is within a first value range, and the second value is within a second value range. The first value range and the second value range do not overlap. The time-to-live field value within the first value range is used to identify the pruned message, and the time-to-live field value within the second value range is used to identify the message that can be pruned.

4. The method according to claim 3, characterized in that, The first value is the upper bound of the first value range.

5. The method according to any one of claims 2 to 4, characterized in that, The method further includes: Received the third message; Obtain the value of the time-to-live field of the third message; In response to the fact that the value of the time-to-live field of the third message is greater than a preset value, the value of the time-to-live field of the third message is reduced by 1 before transmission, wherein the preset value is greater than or equal to the lower bound of the first value range and less than or equal to the upper bound of the first value range.

6. The method according to claim 5, characterized in that, After obtaining the value of the time-to-live field of the third message, the method further includes: In response to the third message's time-to-live field value being equal to the preset value, the third message is discarded.

7. The method according to any one of claims 2 to 4, characterized in that, The method further includes: The fourth message has been received; In response to congestion and packet loss, and if the value of the time-to-live field of the fourth packet is less than the lower bound of the second value range, the fourth packet is discarded.

8. The method according to any one of claims 1 to 4, characterized in that, The step of storing the second message in the transmission queue for transmission includes: The distinguishing service code point of the second message is changed from the first priority value to the second priority value, wherein the priority corresponding to the second priority value is higher than the priority corresponding to the first priority value; The second message is stored in the transmission queue corresponding to the priority value of the second priority value and transmitted.

9. A message identification method, characterized in that, Applied to the receiving end, the method includes: The target message was received; Obtain the value of the time-to-live field of the target message; In response to the value of the time-to-live field of the target message indicating that the target message is a truncated message, the target message is identified as a truncated message.

10. The method according to claim 9, characterized in that, After identifying the target message as a truncated message, the method further includes: In response to having the capability to process trimmed packets, packet trimming feedback processing is performed on the target packet; or, In response to the inability to process the truncated message, the target message is discarded.

11. A message transmission method, characterized in that, include: Obtain the target message to be entered into the network; In response to the target message being a message that can be pruned, the value of the time-to-live field of the target message is set to a second value, wherein the second value is used to indicate that the target message is a message that can be pruned; The target message is transmitted to the network node.

12. The method according to claim 11, characterized in that, The second value is within a second value range, and the second value range does not overlap with the first value range. The time-to-live field value within the first value range is used to identify the clipped message, and the time-to-live field value within the second value range is used to identify the message that can be clipped.

13. An electronic device, characterized in that, The electronic device includes a processor and a memory, the memory storing programs or instructions that can run on the processor, the programs or instructions being executed by the processor to implement the steps of the training method as described in any one of claims 1 to 12.

14. A readable storage medium, characterized in that, The readable storage medium stores a program or instructions that, when executed by a processor, implement the steps of the method as described in any one of claims 1 to 12.

15. A computer program product comprising a computer program stored on a non-transitory computer-readable storage medium, the computer program including program instructions that, when executed by a computer, cause the computer to perform the steps of the method as described in any one of claims 1 to 12.

Images