Message processing method, system, electronic device and computer storage medium
By reserving and sharding the service message flow in the deterministic network according to the scheduling cycle of edge nodes, the problem of periodic queue overflow in the network is solved, and uncontrollable delay and jitter are avoided in the deterministic network. effect.
Patent Information
- Application Number
- PCT/CN2024/073228
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- Priority Date
- 2023-06-20
- Filing Date
- 2024-01-19
- Publication Date
- 2024-12-26
AI Technical Summary
In deterministic networks, because the entire network cycle mapping relationship is not controlled, the bandwidth over-allocation problems caused by ultra-long packets, resulting in uncontrollable delay and jitter.
By obtaining the message characteristics of the service message flow, including service bandwidth requirements and maximum message length, the maximum allowable message length is determined based on the scheduling cycle of the edge node, and when the maximum allowable message length is less than the maximum message length , perform bandwidth reservation deduction processing on the service message flow, and send shard instructions to the edge node. The edge node slices the service message flow according to the shard instructions, and puts shard messages into multiple periodic queues. Processed in the process.
It effectively avoids the periodic queue overflow of nodes, reduces the occurrence of uncontrollable delays and jitter phenomena, and ensures network stability and low latency and low jitter service requirements.
Smart Images

Figure CN2024073228_26122024_PF_FP_ABST
Abstract
Description
Message processing method, system, electronic device and computer storage medium
[0001] Related applications
[0002] This application claims priority to Chinese patent application No. 202310740975.7 filed on June 20, 2023, the entire contents of which are incorporated by reference into this application. Technical Field
[0003] The present application relates to the field of communication technology, and in particular to a message processing method, system, electronic device and computer storage medium. Background Art
[0004] In deterministic periodic scheduling of deterministic networks, the service carrying capacity of each period is fixed. However, since the periodic mapping relationship of the entire network is uncontrolled, there will be bandwidth over-allocation problems caused by ultra-long messages. For example, as shown in Figure 1, the service flows sent from nodes A and E converge at node C and happen to fall into the same scheduling period queue 3 of node C. However, the 3 scheduling period queues of node C are not sufficient to carry these two messages, resulting in uncontrollable delay and jitter. To avoid this defect, the whole packet scheduling method is currently used. However, when encountering low-bandwidth, long messages, this method has a large instantaneous bandwidth occupancy. Because it is a single message, it cannot be adjusted to different queues, resulting in exceeding its scheduling capacity and generating uncontrollable delay and jitter.
[0005] Summary of the Invention
[0006] The main purpose of this application is to provide a message processing method, system, electronic device and computer storage medium, aiming to solve the technical problem of how to avoid the overflow of the periodic queue of nodes in a deterministic network, which leads to the occurrence of uncontrollable delay and jitter.
[0007] To achieve the above objectives, the present application provides a message processing method, which is applied to a controller connected to an edge node in a deterministic network, comprising:
[0008] Acquire message characteristics of the service message flow, wherein the message characteristics include service bandwidth requirements and maximum message length;
[0009] Determining a maximum allowable message length based on the service bandwidth requirement and the scheduling period of the edge node;
[0010] When the maximum allowed message length is less than the maximum message length, bandwidth reservation deduction processing is performed on the service message flow, and a fragmentation instruction is sent to the edge node, wherein the edge node fragments the service message flow according to the fragmentation instruction and processes the fragmented messages according to multiple periodic queues.
[0011] In addition, to achieve the above objectives, the present application provides a message processing method, which is applied to an edge node connected to a controller in a deterministic network, comprising:
[0012] Fragmenting the service message flow according to the received fragmentation instruction to obtain fragmented messages, wherein the fragmentation instruction is an instruction sent by the controller when the maximum allowed send message length is less than the maximum message length of the service message flow, and the maximum allowed send message length is a message length determined according to the service bandwidth requirement of the service message flow;
[0013] The fragment messages are encapsulated, and according to the fragment order of fragment processing, the encapsulated fragment messages are sequentially put into the periodic queue of the edge node for processing.
[0014] In addition, to achieve the above-mentioned purpose, the present application provides a message processing system, including a controller and edge nodes connected to each other in a deterministic network,
[0015] The controller is configured to obtain message characteristics of a service message flow, wherein the message characteristics include a service bandwidth requirement and a maximum message length; determine a maximum allowable send message length based on the service bandwidth requirement and a scheduling period of the edge node; and perform bandwidth reservation deduction processing on the service message flow if the maximum allowable send message length is less than the maximum message length, and send a fragmentation instruction to the edge node;
[0016] The edge node is configured to fragment the service message flow according to the received fragmentation instruction to obtain fragmented messages; encapsulate the fragmented messages, and according to the fragmentation order of the fragmentation processing, put the encapsulated fragmented messages into multiple cycle queues of the edge node for processing.
[0017] In addition, to achieve the above-mentioned purpose, the present application also provides an electronic device, which includes: a memory, a processor, and a message processing program stored on the above-mentioned memory and capable of running on the above-mentioned processor. When the above-mentioned message processing program is executed by the above-mentioned processor, the steps of the message processing method as described above are implemented.
[0018] In addition, to achieve the above-mentioned purpose, the present application also provides a computer storage medium, on which a message processing program is stored. When the message processing program is executed by a processor, the steps of the message processing method as described above are implemented. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] Figure 1 is a schematic diagram of node periodic queue conflicts in a deterministic network;
[0020] Figure 2 is a schematic diagram of node periodic queue scheduling in a deterministic network;
[0021] FIG3 is a schematic diagram of the networking of nodes in a deterministic network;
[0022] FIG4 is a flow chart of a first embodiment of the message processing method of the present application;
[0023] FIG5 is a flow chart of the fragmentation processing of a service message process in the message processing method of the present application;
[0024] FIG6 is a flow chart of a second embodiment of the message processing method of the present application;
[0025] FIG7 is a flow chart of a third embodiment of the message processing method of the present application;
[0026] FIG8 is a schematic diagram of the device structure of the hardware operating environment involved in the message processing method of this application.
[0027] The purpose, features and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION
[0028] It should be understood that the specific embodiments described herein are only used to explain the present application and are not intended to limit the present application.
[0029] Deterministic networking technology is one of the supporting technologies for future 6G communications. Standardization work is currently underway in multiple standards organizations. In packet data communication scenarios, to ensure low latency and low jitter service requirements, different approaches are adopted in different scenarios. In wide and local area networks, a time-based scheduling approach is used. Optionally, the time period can have a fixed duration, or multiple durations can exist, requiring the different scheduling periods to be powers of two, such as 1T, 2T, 4T, or 8T. T represents a duration, for example, T = 10us, where us represents microseconds. Each time period is supported by a series of queues. For example, as shown in Figure 2, port queues Q1, Q2, through Q8 have a scheduling period of 1T. Port queues Q9, Q10, and Q12 have a scheduling period of 2T. Port queues Q13 and Q14 have a scheduling period of 4T. The queue length for each category is different. Calculated based on the port rate R = 10 Gbps and T = 10 μs, the queue lengths for Q1-Q8 are QL = 100,000 bits, the queue lengths for Q9-Q12 are QL = 200,000 bits, and the queue lengths for Q13-Q14 are QL = 400,000 bits.
[0030] For example, as shown in Figure 3, a deterministic network forwarding process consists of a control unit (e.g., an external controller) and five nodes: PE1, PE2, PE3, PE4, and PE5. There are service flows: flow1, flow2, flow3, and flow_n. P1 and interface 1 are configured on P1. During path calculation, paths are arranged based on parameters such as bandwidth. Assume that the port rate R from node P1 to node PE4 is 10 Gbps, and there are 100 service flows with a rate of nearly 100 Mbps. If the service flow is an 8K video stream, with a 1500-byte packet transmitted every 125 μs, the total bandwidth is 8000 * 1500 * 8 = 96,000,000 bps, nearly 100 Mbps. The service flow bandwidth calculation here is based on a time window of seconds. If the bandwidth is scaled to 10 μs, the instantaneous service traffic is very large. If service flow 1 is expected to be 100 Mbps, as shown by the dotted line, and is continuously transmitted at 100 Mbps, in reality, the service sends large (1500-byte) packets every 125 μs. If the interface sending from PE1 to P1 is a 10 Gbps interface, the actual transmission rate is 10 Gbps for approximately 1.2 μs. During this time, no packets are transmitted for this service. Even if rate limiting is applied on PE1's inbound interface, due to Ethernet's full-packet transmission and reception characteristics, PE1 still waits for the complete packet to be received before sending it out the outbound interface. This results in a transient high-burst service flow on the outbound interface. This feature, however, can lead to queue overflow issues in periodic scheduling. For interface 1 of P1, queue Q, with a scheduling period of T = 10µs, can accommodate 100,000 bits (equivalent to 12,500 bytes) within a single period. This means that if nine services similar to Flow 1 arrive simultaneously (12,500 / 1,500 = approximately 8.3), the queue will overflow, despite the expected capacity of nearly 100 service flows. Therefore, preventing periodic queue overflows at nodes in deterministic networks, which can lead to uncontrollable latency and jitter, is crucial.
[0031] This embodiment provides a method for performing fixed-length fragmentation on packets at the edge device's outbound interface based on bandwidth requirements and the number of queues corresponding to the scheduling period. This method also allows for bandwidth expansion to prevent overflow of the periodic queues at the central aggregation node. The controller obtains the scheduling period and number of queues at each node, overlays the interface rate, service bandwidth requirements, and maximum packet length, and comprehensively determines the services requiring fixed-length fragmentation, the fragment length, and the additional bandwidth requirements. The controller then selects a path and issues the fragmentation operation.
[0032] 4 , the present application provides a message processing method, which is applied to a controller connected to an edge node in a deterministic network. In a first embodiment of the message processing method, the message processing method includes:
[0033] Step S10: Acquire message characteristics of the service message flow, wherein the message characteristics include service bandwidth requirements and maximum message length;
[0034] In this embodiment, a controller connected to an edge node in a deterministic network is first started, wherein the controller can be set in a device or a server.
[0035] In one embodiment, the controller can collect topology and node information, such as the node scheduling period, the number of queues corresponding to the node scheduling period, and the port rate. The node information can include edge node information. Edge nodes can be edge access nodes for service message flows or boundary nodes for deterministic forwarding domains. Each node (e.g., an edge node) can be configured with a device for generating or forwarding service message flows.
[0036] In one embodiment, when a controller receives a service establishment or access request sent by another node, it needs to obtain the message characteristics of the established service flow or the accessed service flow based on the service establishment or access request, such as the service flow matching characteristics, service bandwidth requirements, and maximum message length. Among them, the message characteristics of the service message flow can be set as the service flow matching characteristics, service bandwidth requirements, and maximum message length. The service bandwidth requirement can be the service bandwidth required for the service message flow when it flows into the node. The maximum message length can be the maximum message length within the message lengths corresponding to each message in the service message flow. The message length can be expressed in bytes.
[0037] Step S20: determining the maximum allowed message length to be sent based on the service bandwidth requirement and the scheduling period of the edge node;
[0038] In one embodiment, after obtaining the service bandwidth requirements of a service message flow, the controller also needs to determine the scheduling period of the edge node, so as to determine the maximum allowable send message length based on the scheduling period of the edge node. The edge node can be a boundary node in a deterministic network, or a node through which the service message flow accurately passes. The maximum allowable send message length can be the maximum message length that the edge node can send in a single transmission during the scheduling period.
[0039] In one embodiment, Max_Flow_allow_len = BW_Q * T / 8, rounded up. Note that the unit of T here must match and convert to the bandwidth unit. Dividing by 8 converts bytes to bits. T is the scheduling period of the edge node. BW_Q is the service bandwidth requirement. Max_Flow_allow_len is the maximum allowed message length.
[0040] Step S30: When the maximum allowed message length is less than the maximum message length, bandwidth reservation deduction is performed on the service message flow, and a fragmentation instruction is sent to the edge node, wherein the edge node fragments the service message flow according to the fragmentation instruction, and processes the fragmented messages according to multiple periodic queues.
[0041] In one embodiment, if it is detected that the maximum allowable message length is less than the maximum message length, the edge node can determine that the maximum message length allowed to be sent is less than the maximum message length of the service message flow at a certain moment. Continuing to send the service message flow to the edge node at this time will cause the edge node's periodic queue to overflow. Therefore, in this embodiment, to avoid this phenomenon, bandwidth reservation deduction processing can be performed on the service message flow. Specifically, the bandwidth required for the service message flow is reserved in the edge node so that the service message flow can be processed according to the reserved bandwidth after entering the edge node. At the same time, the controller also issues a fragmentation instruction to the edge node. After receiving the fragmentation instruction, the edge node can directly fragment the service message flow and then process the fragmented message segments separately according to the multiple periodic queues in the edge node to avoid periodic queue overflow.
[0042] For example, as shown in Figure 5, after the device and controller are started, the controller collects topology and node information, including the scheduling period, the number of queues corresponding to the scheduling period, and the port rate. Upon receiving a service establishment or access request, the controller obtains service flow matching characteristics (i.e., matching characteristics of the service packet flow), port requirements (BW_Q), and the maximum packet length (Max_Pkt_Len). The controller obtains the service edge access node (which can also be a boundary node in a deterministic forwarding domain, i.e., an edge node) and, based on the node scheduling period T (i.e., the scheduling period of the edge node) and the service bandwidth requirement (BW_Q), calculates the maximum allowable send length (i.e., the maximum allowable send packet length): Max_Flow_allow_len = BW_Q * T / 8, rounded up. Note that the units of T here must match and be converted to bandwidth units, and dividing by 8 is the conversion between bytes and bits. Whether fragmentation is required is determined by whether Max_Flow_allow_len >= Max_Pkt_Len. If it is true, it means that no additional fragmentation processing is required. The bandwidth can be directly reserved in the controller, and the edge device is notified to forward normally. There is no need to issue fragmentation instructions. If it is not true, further judgment is made on whether bandwidth expansion is needed. The judgment basis is whether Max_Flow_allow_len>=(Max_Pkt_Len / Q_num) is established (that is, we want to divide the message into fragments with the same number of queues and put them into each queue. The number of fragments can only be no more than the number of queues Q_num. If the size of the fragments is still longer than the maximum allowed length of the message to be sent according to this fragmentation method, bandwidth expansion is required. If it is not true, bandwidth expansion is required. The controller expands the service bandwidth demand. The expanded bandwidth demand Expand_BW is calculated according to the following formula:
[0043] Expand_BW=BW_Q*((Max_Pkt_Len / Q_num) / Max_Flow_allow_len).
[0044] The controller reserves bandwidth for services and deducts available bandwidth according to Expand_BW. At the same time, it issues fragmentation instructions to edge nodes, with the fragment length being Max_Pkt_Len / Q_num, rounded up. The controller reserves bandwidth for services and deducts available bandwidth according to BW_Q. At the same time, it issues fragmentation instructions to edge nodes, with the fragment length being Max_Flow_allow_len. The edge node matches the relevant service flow and performs fragmentation processing according to the fragment length issued by the controller. It also encapsulates the fragmented messages in a certain way (to facilitate subsequent reassembly) and then places them into the queue being scheduled and subsequent queues in the order of the fragments. The specific message encapsulation is not limited here. It should be noted that service messages are not always sent according to the maximum length. For messages smaller than the maximum packet length, an additional layer of judgment can be added. According to the following relationship between the actual message length Pkt_Len and Max_Flow_allow_len, the number of slices Seg_Num = Pkt_Len / Max_Flow_allow_len, rounded up, is determined. The message is then segmented into Seg_Num segments to avoid encapsulating the message according to the Max_Flow_allow_len length, which would result in the remaining message being too small, requiring padding and wasting bandwidth. If a large amount of additional encapsulation is added, the controller will need to reserve and deduct bandwidth accordingly, which is not described here. If there is a landing edge node, the fragmented message is reassembled and restored based on the encapsulation and outbound interface characteristics, and the reassembled message is forwarded.
[0045] In this embodiment, the maximum allowable message length to be sent is determined in the controller according to the service bandwidth demand of the service message flow and the scheduling cycle of the edge node. When the maximum allowable message length to be sent is less than the maximum message length of the service message flow, bandwidth reservation deduction processing is performed, and a fragmentation instruction is sent to the edge node, so that the edge node fragments the service message flow according to the fragmentation instruction. This can achieve the goal of not requiring full network cycle orchestration, but only requiring bandwidth reservation deduction processing in the controller and fragmentation processing in the edge node, so that multiple cycle queues can process the fragmented messages after fragmentation respectively, thereby avoiding the edge node from receiving a service data flow and receiving a message that is too long instantly, causing the node's cycle queue to overflow, resulting in uncontrollable delay and jitter.
[0046] Based on the first embodiment of the present application, a second embodiment of the message processing method of the present application is proposed. In this embodiment, referring to FIG6 , step S30 performs bandwidth reservation deduction processing on the service message flow and sends a fragmentation instruction to the edge node, including:
[0047] Step a, determining a first ratio between the maximum message length and the number of queues corresponding to the scheduling period;
[0048] In this embodiment, before performing bandwidth reservation deduction processing on the service message flow, it is necessary to first determine whether bandwidth expansion processing is required for the required bandwidth to avoid the phenomenon that the service demand bandwidth is too small, resulting in a small reserved bandwidth and a periodic queue overflow. In one embodiment, the maximum message length and the number of queues corresponding to the scheduling period can be first determined to determine the message length that each scheduling period queue needs to process when the edge node receives the maximum message length at a certain instant. That is, the first ratio between the maximum message length and the number of queues corresponding to the scheduling period is directly calculated, and this first ratio is used as the message length that each scheduling period queue needs to process. For example, if the maximum message length is Max_Pkt_Len and the number of queues corresponding to the scheduling period is Q_num; then the message length is Max_Pkt_Len / Q_num.
[0049] Step b: When the first ratio is less than or equal to the maximum allowed send message length, perform bandwidth reservation deduction processing on the service message flow according to the service bandwidth demand, and send the maximum allowed send message length and fragmentation instruction to the edge node.
[0050] In one embodiment, the first ratio is compared with the maximum allowable send message length. If the first ratio is less than or equal to the maximum allowable send message length, it can be determined that the message length to be processed by each scheduling period queue is less than or equal to the maximum allowable send message length allowed by the edge node. In this case, bandwidth expansion processing is not required, and bandwidth reservation deduction processing is directly performed on the service message flow based on the service bandwidth demand. In other words, the edge node reserves the required bandwidth corresponding to the service bandwidth demand. The maximum allowable send message length and a fragmentation instruction are then sent to the edge node. Upon receiving the fragmentation instruction and the maximum allowable send message length, the edge node can perform fragmentation processing to avoid periodic queue overflow.
[0051] For example, as shown in Table 1, if the service message flow is Flow 1, and the service message flow has matching characteristics, service bandwidth requirements, maximum message length, edge access node (i.e., edge node), scheduling period (i.e., scheduling period of the edge node), and number of queues corresponding to the period (i.e., number of queues corresponding to the scheduling period), Table 1 can be used to determine whether the service message flow needs to be fragmented, and if so, the fragment length, whether bandwidth expansion is required, and the expanded bandwidth obtained after the bandwidth expansion.
[0052] Table 1
[0053] In one embodiment, after the device and controller are started, the controller collects topology and node information, such as the periodic scheduling length (i.e., the scheduling period in Table 1), the number of queues corresponding to the scheduling period, and the port rate. Upon receiving a request to establish or access service flow Flow1, the controller obtains matching characteristics, service bandwidth requirements, and maximum packet length, as described in Table 1 above. The controller also obtains the edge node, PE1, and determines the scheduling period and the number of queues corresponding to the period based on the service characteristics. Based on the scheduling period and service bandwidth requirements, the controller calculates the maximum allowable transmit packet length. For example, if the maximum allowable transmit packet length is 250 bytes, the controller determines whether fragmentation is necessary. If the maximum allowable transmit packet length is 250 bytes, which is less than the maximum packet length of 512 bytes, fragmentation is required, and then determines whether bandwidth expansion is necessary. In one embodiment, if the maximum allowable transmit packet length is 250 bytes, the first ratio between the maximum packet length and the number of queues corresponding to the scheduling period is 512 / 8 = 64. Since 250 is greater than 64, it is determined that bandwidth expansion is not necessary, as shown in Table 2 below.
[0054] Table 2
[0055] In one embodiment, the controller reserves bandwidth and deducts available bandwidth according to the service bandwidth demand. At the same time, it issues a fragmentation instruction to the edge node, and the maximum allowed fragment length is 250. In one embodiment, when the edge node matches the service message flow Flow1, it performs fragmentation processing according to the fragment length issued by the controller, and encapsulates the fragmented message obtained after fragmentation, and then puts it into the scheduled periodic queue or subsequent periodic queue in the order of fragmentation. Among them, the original message corresponding to the maximum message length is rounded up according to Seg_Num = Pkt_Len / Max_Flow_allow_len, and Seg_Num = 512 / 250 is obtained. After rounding up, Seg_Num = 3, the message length after slicing is seg_pkt_len = 512 / 3 = 170 bytes, and fragmentation is performed according to 170 bytes. Among them, Seg_Num is the number of slices. The three fragmented messages are sequentially placed into the corresponding periodic queues, as shown in Figure 4 , including periodic queues Q1, Q2, Q3, Q4, Q5, Q6, Q7, and Q8. If the currently scheduled periodic queue is Q4 and there is free space, the first fragmented message is placed there, followed by the second fragmented message in Q5 and the third fragmented message in Q6. The position of a message in a queue is not fixed, depending on the number of existing messages within it. In one embodiment, if node PE4 is a deterministic domain boundary, the fragmented messages are reassembled and restored at node PE4, and the restored message is sent.
[0056] In this embodiment, by determining a first ratio between the maximum message length of the service message flow and the number of queues corresponding to the scheduling period, and when the first ratio is less than or equal to the maximum allowed message length, the service message flow is subjected to bandwidth reservation deduction processing according to the service bandwidth demand, and the maximum allowed message length and fragmentation instruction are sent to the edge node so that the edge node can fragment the service message flow to avoid the occurrence of periodic queue overflow.
[0057] After determining a first ratio between the maximum message length and the number of queues corresponding to the scheduling period, the method further includes:
[0058] Step c: when the first ratio is greater than the maximum allowable send message length, expanding the service bandwidth demand to obtain an expanded bandwidth demand;
[0059] In this embodiment, if the first ratio is greater than the maximum allowable sendable message length, it can be determined that the message length required to be processed by each scheduling period queue exceeds the maximum allowable sendable message length. If the service bandwidth demand is not expanded, the period queue may overflow. Therefore, in this embodiment, the service bandwidth demand can be expanded to obtain a corresponding expanded bandwidth demand.
[0060] For example, if Max_Flow_allow_len>=(Max_Pkt_Len / Q_num), it is determined that the service bandwidth demand needs to be expanded. That is, the messages in the service message flow are divided into slices with the same number of queues (wherein the size of the slices divided can be the first ratio obtained by the above calculation), and placed in each queue, and the slices that can be divided can only be no more than the number of slices in the queue Q_num. However, if the slices divided are larger than the maximum allowed length of the message to be sent, expansion processing is required. In one embodiment, the controller performs a service bandwidth demand expansion operation to obtain an expanded bandwidth demand, if the expanded bandwidth demand is Expand_BW. The method for determining the expanded bandwidth demand can be determined in the following manner. That is:
[0061] Expand_BW=BW_Q*((Max_Pkt_Len / Q_num) / Max_Flow_allow_len).
[0062] The maximum packet length is Max_Pkt_Len, the number of queues corresponding to the scheduling period is Q_num, Max_Flow_allow_len is the maximum allowed packet length, and BW_Q is the service bandwidth requirement.
[0063] Step d: performing bandwidth reservation deduction processing on the service message flow according to the expanded bandwidth demand, rounding up the first ratio to obtain a target fragment length, and sending the target fragment length and fragmentation instruction to the edge node.
[0064] In one embodiment, after determining the expanded bandwidth demand, the controller can directly perform bandwidth reservation deduction processing according to the expanded bandwidth demand, that is, reserve the bandwidth occupied by the demand corresponding to the expanded bandwidth demand at the edge node. Then, the target fragment length is determined, and the target fragment length and fragment instruction are sent to the edge node together. After receiving the fragment instruction and target fragment length, the edge node can perform fragment processing to avoid the occurrence of periodic queue overflow. In one embodiment, when determining the target fragment length, the calculated first ratio can be directly rounded up to obtain the target fragment length. For example, if the calculated first ratio is 1699.4 bytes, the target fragment length can be directly determined to be 1700 bytes.
[0065] For example, as shown in Table 3, if the service message flow is Flow 2, and the service message flow has matching characteristics, service bandwidth requirements, maximum message length, edge access node (i.e., edge node), scheduling period (i.e., scheduling period of the edge node), and number of queues corresponding to the period (i.e., number of queues corresponding to the scheduling period), Table 2 can be used to determine whether the service message flow needs to be fragmented, and if so, the fragment length, whether bandwidth expansion is required, and the expanded bandwidth obtained after the bandwidth expansion.
[0066] Table 3
[0067] In one embodiment, after the device and controller are started, the controller collects topology and node information, such as the periodic scheduling length (i.e., the scheduling period in Table 3), the number of queues corresponding to the scheduling period, and the port rate. Upon receiving a request to establish or access service message flow Flow2, the controller obtains matching characteristics, service bandwidth requirements, and maximum message length, as described in Table 3 above. The controller also obtains the edge node, PE1, and determines the scheduling period and the number of queues corresponding to the period based on the service characteristics. Based on the scheduling period and service bandwidth requirements, the maximum allowable message length is calculated. For example, if the maximum allowable message length is 250 bytes, then if the maximum allowable message length of 250 bytes is less than the maximum message length of 1500 bytes, then fragmentation is determined to be necessary. A determination is then made as to whether bandwidth expansion is necessary. In one embodiment, if the maximum allowable message length is 250, then the first ratio between the maximum message length and the number of queues corresponding to the scheduling period is 1500 / 4 = 375. Since 250 is less than 375, it is determined that bandwidth expansion is necessary, as shown in Table 4 below.
[0068] Table 4
[0069] In one embodiment, the controller reserves bandwidth and deducts available bandwidth according to the service bandwidth demand. At the same time, it issues a fragmentation instruction to the edge node, and the fragmentation length = 1500 / 4 = 375 bytes. In one embodiment, when the edge node matches the service message flow Flow2, it performs fragmentation processing according to the fragmentation length issued by the controller, and encapsulates the fragmented message obtained after fragmentation, and then puts it into the scheduled periodic queue or subsequent periodic queue according to the order of fragmentation. Among them, the original message corresponding to 1500 bytes is fragmented according to 375 bytes. In one embodiment, if there is a node PE4 as a deterministic domain boundary, the fragmented message is reassembled and restored in the node PE4, and the restored message is sent out.
[0070] In this embodiment, when a first ratio between the maximum message length of a service message flow and the number of queues corresponding to a scheduling period is greater than a maximum allowable send message length, the service bandwidth demand is expanded to obtain an expanded bandwidth demand. Then, bandwidth reservation deduction processing is performed on the service message flow according to the expanded bandwidth demand. The target fragment length and fragmentation instruction determined according to the first ratio are sent to the edge node so that the edge node can fragment the service message flow to avoid the occurrence of period queue overflow.
[0071] 7 , the present application provides a message processing method, which is applied to an edge node connected to a controller in a deterministic network. In a third embodiment of the message processing method, the message processing method includes:
[0072] Step S100, fragmenting the service message flow according to the received fragmentation instruction to obtain fragmented messages, wherein the fragmentation instruction is an instruction sent by the controller when the maximum allowable sent message length is less than the maximum message length of the service message flow, and the maximum allowable sent message length is a message length determined according to the service bandwidth requirement of the service message flow;
[0073] In this embodiment, the controller and the edge node may be the controller and the edge node in the first or second embodiment described above.
[0074] In one embodiment, the controller can obtain the service bandwidth requirement and maximum message length of the service message flow and use them as message characteristics of the service message flow. Then, the maximum allowable send message length is determined based on the service bandwidth requirement and the scheduling period of the edge node. And when the maximum allowable send message length is less than the maximum message length, bandwidth reservation deduction processing is performed and a fragmentation instruction is sent to the edge node. In one embodiment, performing bandwidth reservation deduction processing and sending fragmentation instructions to the edge node includes: determining a first ratio between the maximum message length and the number of queues corresponding to the scheduling period, when the first ratio is less than or equal to the maximum allowable send message length, performing bandwidth reservation deduction processing on the service message flow according to the service bandwidth requirement, and sending the maximum allowable send message length and fragmentation instructions to the edge node. In one embodiment, when the first ratio is greater than the maximum allowable send message length, the service bandwidth requirement is expanded to obtain an expanded bandwidth requirement, bandwidth reservation deduction processing is performed on the service message flow according to the expanded bandwidth requirement, and the first ratio is rounded up to obtain a target fragmentation length.
[0075] In one embodiment, after the edge node matches the corresponding service message flow and receives the fragmentation instruction sent by the controller, it can directly fragment the messages in the service message flow according to the fragmentation instruction to obtain fragmented messages.
[0076] Step S200: encapsulate the fragmented messages, and put the encapsulated fragmented messages into multiple periodic queues of the edge node in sequence for processing according to the fragmentation order of fragmentation processing.
[0077] In one embodiment, after fragmenting a service message flow to obtain individual fragmented messages, the edge node may encapsulate the fragmented messages and sequentially place the encapsulated fragmented messages into multiple periodic queues within the edge node for processing, such as the currently scheduled periodic queue and subsequent periodic queues. The fragmentation order of the fragmentation processing may be determined based on the position of each fragmented message within the service message flow.
[0078] In this embodiment, the service message flow is fragmented in the edge node according to the received fragmentation instruction to obtain fragmented messages, the fragmented messages are encapsulated, and the encapsulated fragmented messages are placed in sequence into multiple periodic queues for processing in the fragmentation order, so that fragmentation processing can be realized in the edge node, and multiple periodic queues can process the fragmented messages after fragmentation respectively, avoiding the edge node from receiving the service data flow. The message is too long at the moment, causing the node's periodic queue to overflow, resulting in uncontrollable delay and jitter.
[0079] In one embodiment, fragmenting the service message flow according to the received fragmentation instruction to obtain fragmented messages includes:
[0080] Step e: determining the received fragmentation instruction and target fragmentation length, and fragmenting the service message flow according to the target fragmentation length based on the fragmentation instruction to obtain fragmented messages.
[0081] In this embodiment, after receiving the fragmentation instruction and target fragmentation length sent by the controller, the edge node can directly fragment the service message flow according to the fragmentation instruction, and the message corresponding to the maximum message length can be fragmented according to the target fragmentation length to obtain fragmented messages.
[0082] In this embodiment, the service message flow is fragmented according to the target fragment length according to the fragmentation instruction to obtain fragmented messages, thereby ensuring the validity of the obtained fragmented messages.
[0083] The service message flow is fragmented according to the received fragmentation instruction to obtain fragmented messages, which also includes:
[0084] Step f, determining the received fragmentation instruction and the maximum allowed length of the message to be sent, and based on the fragmentation instruction, fragmenting the service message flow according to the maximum allowed length of the message to be sent to obtain fragmented messages.
[0085] In this embodiment, after receiving the fragmentation instruction and the maximum allowed message length sent by the controller, the edge node can directly fragment the business message flow according to the fragmentation instruction, and the message corresponding to the maximum message length can be fragmented according to the maximum allowed message length to obtain a fragmented message.
[0086] In this embodiment, the service message flow is fragmented according to the maximum allowed message length according to the fragmentation instruction to obtain fragmented messages, thereby ensuring the validity of the obtained fragmented messages.
[0087] The service message flow is fragmented according to the maximum allowed message length to obtain fragmented messages, including:
[0088] Step g, determining an actual message length corresponding to the service message flow, and rounding up a second ratio between the actual message length and the maximum allowed send message length to obtain a number of slices;
[0089] Step h: fragment the service message flow according to the number of slices to obtain fragmented messages.
[0090] In this embodiment, when slicing a service packet flow, the actual packet length corresponding to each packet in the service packet flow, such as the maximum packet length, can also be determined. The second ratio between the actual packet length of the edge node and the maximum allowed transmit packet length is then rounded up to obtain the number of slices. For example, Seg_Num = Pkt_Len / Max_Flow_allow_len, rounded up. The actual packet length can be the length, in bytes, of each packet in the service packet flow.
[0091] In this embodiment, the number of slices is determined according to the actual message length corresponding to the service message flow, and then fragmentation processing is performed according to the number of slices to obtain fragmented messages, thereby ensuring the validity of the obtained fragmented messages.
[0092] In addition, the present application also provides an electronic device, which includes a memory, a processor, and a message processing program stored in the memory and runnable on the processor. When the message processing program is executed by the processor, the steps of the message processing method described above are implemented.
[0093] In addition, in one embodiment, FIG8 is a schematic diagram of the structure of an electronic device according to an embodiment of the present application. As shown in FIG8 , at the hardware level, the electronic device includes a processor and optionally an internal bus, a network interface, and a memory. The memory may include internal memory, such as high-speed random-access memory (RAM), and may also include non-volatile memory, such as at least one disk storage device. Of course, the electronic device may also include hardware required for other services. The processor, network interface, and memory may be interconnected via an internal bus, which may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, or an EISA (Extended Industry Standard Architecture) bus. Buses can be divided into address buses, data buses, control buses, etc. For ease of illustration, the figure uses only one bidirectional arrow, but this does not mean that there is only one bus or only one type of bus. The memory is used to store programs. Programs may include program code, which includes computer operating instructions. The processor reads the corresponding computer program from the non-volatile memory and then runs it in the memory, forming a shared resource access control device at the logical level. The processor executes the program stored in the memory and is specifically used to perform the steps of the above-mentioned message processing method.
[0094] The specific implementation of the electronic device of the present application is basically the same as the embodiments of the above-mentioned message processing method, and will not be repeated here.
[0095] In addition, to achieve the above-mentioned purpose, the present application also provides a computer storage medium, including a computer-readable storage medium, on which a message processing program is stored. When the message processing program is executed by the processor, the steps of the message processing method as described above are implemented.
[0096] The specific implementation of the computer-readable storage medium of the present application is basically the same as the embodiments of the above-mentioned message processing method, and will not be repeated here.
[0097] It will be appreciated by those skilled in the art that all or some of the steps, systems, and functional modules / units in the methods disclosed above may be implemented as software, firmware, hardware, and appropriate combinations thereof. In hardware implementations, the division between the functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed by several physical components in cooperation. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, a digital signal processor, or a microprocessor, or implemented as hardware, or implemented as an integrated circuit, such as an application-specific integrated circuit. Such software may be distributed on a computer-readable medium, which may include a computer storage medium (or non-transitory medium) and a communication medium (or temporary medium). As is well known to those skilled in the art, the term computer-readable storage medium includes volatile and non-volatile, removable, and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer-readable storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical disk storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to store the desired information and can be accessed by a computer. In addition, it is well known to those skilled in the art that communication media generally contain computer-readable instructions, data structures, program modules, or other data in a modulated data signal such as a carrier wave or other transport mechanism, and may include any information delivery media.
[0098] The above are merely optional embodiments of the present application and do not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.
Claims
1. A message processing method, applied to a controller connected to an edge node in a deterministic network, comprising: Acquire message characteristics of the service message flow, wherein the message characteristics include service bandwidth requirements and maximum message length; Determine the maximum allowed length of the sent message according to the service bandwidth requirement and the scheduling period of the edge node; When the maximum allowed sent message length is less than the maximum message length, bandwidth reservation deduction processing is performed on the service message flow, and a fragmentation instruction is sent to the edge node, wherein the edge node fragments the service message flow according to the fragmentation instruction, and processes the fragmented messages after fragmentation according to multiple periodic queues.
2. The message processing method according to claim 1, wherein: The performing bandwidth reservation deduction processing on the service message flow and sending a fragmentation instruction to the edge node includes: Determine a first ratio between the maximum message length and the number of queues corresponding to the scheduling period; When the first ratio is less than or equal to the maximum allowed send message length, bandwidth reservation deduction processing is performed on the service message flow according to the service bandwidth demand, and the maximum allowed send message length and fragmentation instruction are sent to the edge node.
3. The message processing method according to claim 2, wherein: After determining the first ratio between the maximum message length and the number of queues corresponding to the scheduling period, the method further comprises: When the first ratio is greater than the maximum allowed length of the message to be sent, the service bandwidth requirement is expanded to obtain an expanded bandwidth requirement; The service message flow is subjected to bandwidth reservation deduction processing according to the expanded bandwidth demand, and the first ratio is rounded up to obtain a target fragmentation length, and the target fragmentation length and fragmentation instruction are sent to the edge node.
4. A message processing method, applied to an edge node connected to a controller in a deterministic network, comprising: Slice the service message flow according to the received fragmentation instruction to obtain fragmented messages, wherein the fragmentation instruction is an instruction sent by the controller when the maximum allowed message length is less than the maximum message length of the service message flow, and the maximum allowed message length is a message length determined according to the service bandwidth requirement of the service message flow; The fragment messages are encapsulated, and according to the fragment order of fragment processing, the encapsulated fragment messages are sequentially put into multiple periodic queues of the edge node for processing.
5. The message processing method according to claim 4, wherein: The process of fragmenting the service message flow according to the received fragmentation instruction to obtain fragmented messages includes: Determine the received fragmentation instruction and target fragmentation length, and fragment the service message flow according to the target fragmentation length based on the fragmentation instruction to obtain fragmented messages.
6. The message processing method according to claim 4, wherein: The method of performing fragmentation processing on the service message flow according to the received fragmentation instruction to obtain fragmented messages also includes: Determine the received fragmentation instruction and the maximum allowed length of the sent message, and according to the fragmentation instruction, fragment the service message flow according to the maximum allowed length of the sent message to obtain fragmented messages.
7. The message processing method according to claim 6, wherein: The process of fragmenting the service message flow according to the maximum allowed message length to obtain fragmented messages includes: Determine an actual message length corresponding to the service message flow, and round up a second ratio between the actual message length and the maximum allowed message length to obtain a number of slices; The service message flow is fragmented according to the number of slices to obtain fragmented messages.
8. A message processing system, wherein: The system includes a controller and edge nodes connected to each other in a deterministic network, The controller is configured to obtain message features of a service message flow, wherein the message features include a service bandwidth requirement and a maximum message length; determine a maximum allowed message length to be sent according to the service bandwidth requirement and a scheduling period of the edge node; and perform bandwidth reservation deduction processing on the service message flow when the maximum allowed message length to be sent is less than the maximum message length. And send a sharding instruction to the edge node; The edge node is configured to perform fragment processing on the service message flow according to the received fragmentation instruction to obtain fragmented messages; encapsulate the fragmented messages, and according to the fragmentation order of the fragmentation processing, sequentially put the encapsulated fragmented messages into multiple cycle queues of the edge node for processing.
9. An electronic device, wherein: The electronic device comprises: a memory, a processor, and a message processing program stored in the memory and executable on the processor, wherein the message processing program implements the steps of the message processing method according to any one of claims 1 to 7 when executed by the processor.
10. A computer storage medium, wherein: The computer storage medium stores a message processing program, and when the message processing program is executed by the processor, the steps of the message processing method according to any one of claims 1 to 7 are implemented.
Citation Information
Patent Citations
Method and device for updating mapping relation
CN112825510A
Method and device for realizing whole network period label planning of wide area deterministic network
CN114338555A
Communication method and device
CN116015574A
Method for preventing queue overflow in periodic scheduling, controller and computer readable storage medium
CN116527598A
Message forwarding method and apparatus, and electronic device
WO2022109986A1
