[go: up one dir, main page]

WO2025035627A1 - 缓存分配方法、装置以及设备 - Google Patents

缓存分配方法、装置以及设备 Download PDF

Info

Publication number
WO2025035627A1
WO2025035627A1 PCT/CN2023/131355 CN2023131355W WO2025035627A1 WO 2025035627 A1 WO2025035627 A1 WO 2025035627A1 CN 2023131355 W CN2023131355 W CN 2023131355W WO 2025035627 A1 WO2025035627 A1 WO 2025035627A1
Authority
WO
WIPO (PCT)
Prior art keywords
cache
traffic
network device
value
cache allocation
Prior art date
Application number
PCT/CN2023/131355
Other languages
English (en)
French (fr)
Inventor
宋胜安
周建二
李伟超
黄勇
陶小峰
Original Assignee
鹏城实验室
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 鹏城实验室 filed Critical 鹏城实验室
Publication of WO2025035627A1 publication Critical patent/WO2025035627A1/zh

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/27—Evaluation or update of window size, e.g. using information derived from acknowledged [ACK] packets
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/29—Flow control; Congestion control using a combination of thresholds
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/10—Flow control; Congestion control
    • H04L47/30—Flow control; Congestion control in combination with information about buffer occupancy at either end or at transit nodes
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/70—Admission control; Resource allocation

Definitions

  • the present application relates to the technical field of network transmission management, and in particular to a cache allocation method, device and equipment.
  • Each edge network device is connected to multiple network devices. Since the bandwidth resources available to edge network devices are limited and each congestion control algorithm has different capabilities for preempting bandwidth resources, network devices that use congestion control algorithms with strong bandwidth preemption capabilities will occupy more bandwidth resources, while network devices that use congestion control algorithms with weak bandwidth preemption capabilities will occupy less bandwidth resources. As a result, bandwidth resources cannot be reasonably allocated according to demand, reducing the fairness of bandwidth preemption of different congestion control algorithms.
  • the main purpose of the present application is to provide a cache allocation method, device and equipment, aiming to solve the technical problem in the prior art that bandwidth resources cannot be reasonably allocated according to demand, thereby reducing the fairness of bandwidth preemption of different congestion control algorithms.
  • the present application provides a cache allocation method, the cache allocation method comprising:
  • the traffic of the network device is allocated to a cache based on the cache allocation value.
  • the step before the step of determining the cache allocation value of the network device based on the congestion control algorithm, the occupied bandwidth information and the cache allocation rule, the step further includes:
  • Determining the device type of the network device based on a preset classification cache threshold, the cache amount, and the egress bandwidth;
  • a cache allocation rule is determined.
  • the device types include Cubic algorithm dominant devices and BBR algorithm dominant devices;
  • the step of determining the device type of the network device based on the preset classification cache threshold, the cache amount and the egress bandwidth includes:
  • the network device is determined to be the BBR algorithm leading device.
  • the cache allocation rule includes a scheduling cache rule
  • the congestion control algorithm includes a BBR algorithm
  • the step of determining the cache allocation value of the network device based on the congestion control algorithm, the occupied bandwidth information and the cache allocation rule includes:
  • the device type of the network device is the BBR algorithm dominant device, the total cache value of the port connected to the network device, the egress bandwidth, the BBR occupied bandwidth of the BBR traffic at the current moment, the number of other traffic items that use congestion control algorithms other than the BBR algorithm to transmit traffic, and the number of BBR flows of the BBR traffic are filtered out from the occupied bandwidth information;
  • the BBR traffic is the traffic transmitted using the BBR algorithm
  • the preset scheduling threshold belongs to the cache allocation value.
  • the cache allocation rule further includes a cache allocation value rule
  • the congestion control algorithm further includes a Cubic algorithm
  • the step of determining the cache allocation value of the network device based on the congestion control algorithm, the occupied bandwidth information and the cache allocation rule includes:
  • the device type of the network device is a Cubic algorithm dominant device
  • the total port cache value, the egress bandwidth, the Cubic occupied bandwidth occupied by the Cubic traffic at the current moment, the number of Cubic flows of the Cubic traffic at the current moment, and the number of irrelevant flows that use the Cubic algorithm to transmit traffic are filtered out from the occupied bandwidth information;
  • the Cubic traffic is the traffic transmitted based on the Cubic algorithm
  • the cache value belongs to the cache allocation value.
  • the step of identifying a congestion control algorithm used by a network device when sending traffic includes:
  • the window trend characteristics are compared with different types of preset theoretical window trends to determine a congestion control algorithm used to send the traffic.
  • the step of determining the cache allocation value of the network device based on the congestion control algorithm, the occupied bandwidth information and the cache allocation rule further includes:
  • the total port cache value, the egress bandwidth, the Cubic occupied bandwidth, the number of irrelevant flows, the number of Cubic flows, the number of other flows and the number of BBR flows are filtered out from the flow information table.
  • the cache allocation value of the network device is determined based on the total port cache value, the egress bandwidth, the Cubic occupied bandwidth, the number of irrelevant flows, the number of Cubic flows, the number of remaining flows, the number of BBR flows and the cache allocation rule.
  • the present application also provides a cache allocation device, the cache allocation device comprising:
  • An information collection module is used to identify the congestion control algorithm used by different network devices when sending traffic, and to collect statistics on the bandwidth occupied by the traffic;
  • a determination module configured to determine a cache allocation value of the network device based on the congestion control algorithm, the occupied bandwidth information and a cache allocation rule
  • An allocation module is used to allocate the traffic of the network device to a cache based on the cache allocation value.
  • the determination module is also used to obtain the cache capacity of the traffic cache area in the network device and the export bandwidth of each output port in the network device; determine the device type of the network device based on a preset classification cache threshold, the cache capacity and the export bandwidth; and determine the cache allocation rules based on the device type.
  • the present application also proposes a cache allocation device, which includes: a memory, a processor, and a cache allocation program stored in the memory and executable on the processor, wherein the cache allocation program is configured to implement the steps of the cache allocation method described above.
  • the present application also proposes a storage medium, on which a cache allocation program is stored, and when the cache allocation program is executed by a processor, the steps of the cache allocation method described above are implemented.
  • the present application provides a cache allocation method, apparatus and device. Compared with the prior art in which each congestion control algorithm has different preemption capabilities for bandwidth resources, bandwidth resources cannot be reasonably allocated according to demand, and the fairness of bandwidth preemption of different congestion control algorithms is reduced, in the present application, the congestion control algorithms used by different network devices when sending traffic are identified, and the occupied bandwidth information of the traffic is counted; the cache allocation value of the network device is determined based on the congestion control algorithm, the occupied bandwidth information and the cache allocation rule; and the traffic of the network device is allocated to cache based on the cache allocation value.
  • the cache allocation value that needs to be allocated to the network device is determined according to the congestion control algorithm, the occupied bandwidth information and the cache allocation rule, so that the network device can accurately allocate the network traffic to the cache according to the cache allocation value. That is, in the present application, the cache allocation value that needs to be allocated to the network device is determined according to the congestion control algorithm, the occupied bandwidth information and the cache allocation rule, and the traffic of the network device is allocated to the cache according to the cache allocation value, so that the bandwidth resources can be reasonably allocated according to the demand, thereby improving the fairness of bandwidth preemption of different congestion control algorithms.
  • FIG1 is a schematic diagram of a device structure of a hardware operating environment involved in an embodiment of the present application
  • FIG2 is a schematic diagram of a flow chart of a first embodiment of a cache allocation method of the present application
  • FIG3 is a schematic diagram of the actual working process of the cache allocation method of the present application.
  • FIG4 is a flow chart of a second embodiment of the cache allocation method of the present application.
  • FIG5 is a schematic diagram of a device classification process of network devices in the cache allocation method of the present application.
  • FIG6 is a schematic diagram of a flow chart of a third embodiment of a cache allocation method of the present application.
  • FIG. 7 is a schematic diagram of the structural configuration of the cache allocation device of the present application.
  • FIG. 1 is a schematic diagram of the structure of a cache allocation device in a hardware operating environment involved in an embodiment of the present application.
  • the cache allocation device may include: a processor 1001, such as a central processing unit (CPU), a communication bus 1002, a user interface 1003, a network interface 1004, and a memory 1005.
  • the communication bus 1002 is used to realize the connection and communication between these components.
  • the user interface 1003 may include a display screen (Display), an input unit such as a keyboard (Keyboard), and the user interface 1003 may also include a standard wired interface and a wireless interface.
  • the network interface 1004 may include a standard wired interface and a wireless interface (such as a wireless fidelity (WIreless-FIdelity, WI-FI) interface).
  • the memory 1005 may be a high-speed random access memory (Random Access Memory, RAM) memory, or a stable non-volatile memory (Non-Volatile Memory, NVM), such as a disk memory.
  • RAM Random Access Memory
  • NVM Non-Volatile Memory
  • the memory 1005 may also be a storage device independent of the aforementioned processor 1001.
  • FIG. 1 does not constitute a limitation on the cache allocation device, and may include more or fewer components than shown in the figure, or a combination of certain components, or a different arrangement of components.
  • the memory 1005 as a storage medium may include an operating system, a data storage module, a network communication module, a user interface module, and a cache allocation program.
  • the network interface 1004 is mainly used for data communication with other devices;
  • the user interface 1003 is mainly used for data interaction with the user;
  • the processor 1001 and the memory 1005 in the cache allocation device of the present application can be set in the cache allocation device, and the cache allocation device calls the cache allocation program stored in the memory 1005 through the processor 1001, and executes the cache allocation method provided in the embodiment of the present application.
  • FIG. 2 is a flow chart of a first embodiment of a cache allocation method of the present application.
  • the executor of this embodiment may be the cache allocation device, which may be an edge network device of electronic devices such as personal computers, smart phones, and tablet computers, or other devices that can achieve the same or similar functions. This embodiment does not limit this. In this embodiment and the following embodiments, the cache allocation method of this application is described using the cache allocation device as an example.
  • the network bandwidth allocation device also includes a traffic identification and classification module (not shown in the figure), a traffic bandwidth statistics module (not shown in the figure), a device cache calculation management module (not shown in the figure), and may also include a flow information table (refer to Figure 3).
  • the traffic identification and classification module is used to identify network devices and the congestion control algorithm used when sending traffic, and classify and record the traffic according to the congestion control algorithm. It can also determine the key value of each group of traffic through the hash algorithm based on the traffic quintuple information; the traffic bandwidth statistics module can monitor the occupied bandwidth information of various traffic in real time; the flow information table is used to record the traffic information obtained by the traffic identification and classification module and the occupied bandwidth information obtained by the traffic bandwidth statistics module, and is updated in real time based on the traffic information and occupied bandwidth information; the device cache calculation management module is used to perform cache calculation and allocation according to the traffic information table, and send it to the data level for execution.
  • the cache allocation method includes:
  • Step S10 identifying the congestion control algorithm used by the network device when sending traffic, and counting the occupied bandwidth information of the traffic;
  • the network device may be a computer, server, mobile phone, tablet or other electronic device that can use the network to transmit data, without specific limitation.
  • the congestion control algorithm can be an algorithm used to adjust the number of data packets sent continuously at a single time by the Transmission Control Protocol (TCP).
  • TCP Transmission Control Protocol
  • the network bandwidth can be gradually adjusted by increasing or decreasing the number of data packets sent at a single time to make it close to the current network's carrying capacity.
  • the congestion control algorithm may include a Cubic algorithm, a BBR (Bottleneck Bandwidth and Round-trip time) algorithm, etc., without specific limitation.
  • congestion control traffic can be regarded as the traffic transmitted by which congestion control algorithm is used, and the traffic is regarded as the congestion control algorithm.
  • traffic transmitted using the Cubic algorithm can be considered Cubic traffic
  • traffic transmitted using the BBR algorithm can be considered BBR traffic.
  • the traffic identification and classification module is used to determine the window trend characteristics of the traffic sent by the network device. According to the window trend characteristics, the congestion control algorithm used by the network device when transmitting traffic is identified, and the traffic bandwidth statistics module is used to count the traffic bandwidth occupied by each network device, that is, the occupied bandwidth information of each congestion control traffic.
  • step S10 in this embodiment further includes:
  • Step S11 extracting window trend characteristics of traffic sent by each network device
  • Step S12 comparing the window trend characteristics with different types of preset theoretical window trends to determine the congestion control algorithm used to send the traffic.
  • the window can be regarded as the congestion window where traffic congestion occurs; the window trend can be regarded as the congestion trend at the congestion window during traffic transmission; and the preset theoretical window trend can be the congestion trend of the congestion window traffic obtained through experimental data.
  • the traffic identification and classification module is first used to collect experimental data, and the Cubic window trend of the Cubic algorithm and the BBR window trend of the BBR algorithm are sorted out through the experimental data, and the window trend characteristics of the network device sending traffic are extracted.
  • the window trend characteristics are compared with the preset theoretical window trends to determine the congestion control algorithm used for sending traffic. That is, the congestion trend at the congestion window of each congestion control algorithm is determined through experimental data, and the congestion trend is used to determine the congestion control algorithm used by the network device to send traffic, which avoids complex calculation processes and improves the efficiency of identifying the congestion control algorithm for sending flows.
  • Step S20 determining a cache allocation value of the network device based on the congestion control algorithm, the occupied bandwidth information and the cache allocation rule;
  • the congestion control algorithm corresponding to each group of traffic is determined, and the maximum bandwidth that can be used by each network device is determined based on the occupied bandwidth information and bandwidth allocation rules of each group.
  • the cache allocation value of each network device is determined based on the maximum bandwidth, so that network devices using each congestion control algorithm can obtain bandwidth resources that can meet their own needs, thereby avoiding a certain congestion control algorithm occupying too much bandwidth resources, causing network devices using other congestion control algorithms to be unable to use bandwidth or too little available bandwidth to send traffic normally, thereby improving the fairness of bandwidth preemption by different congestion control algorithms.
  • the cache calculation management module is used to determine the occupied bandwidth corresponding to each congestion control algorithm at the current moment, and the total cache size of each network device port that can cache traffic, the export bandwidth of each network device port, the total bandwidth occupied by each congestion control algorithm at the current moment and the number of flows of each traffic are analyzed from the occupied bandwidth information.
  • the cache allocation rule is used to determine the cache allocation value of each network device in the cache calculation management module to reasonably allocate bandwidth resources according to the cache allocation value.
  • the egress bandwidth of a network device port may be the maximum bandwidth that the network device can occupy when sending traffic. If the bandwidth is exceeded, port congestion may occur, and packet loss may occur due to congestion.
  • Step S30 Allocate traffic to cache of the network device based on the cache allocation value.
  • the cache allocation value may be a zero value, that is, there is no need to allocate bandwidth to the network device or adjust the traffic cache area of the network device.
  • the bandwidth occupancy of the network devices can be detected in real time during the allocation process until the occupied bandwidth of the network device is equal to the average bandwidth of the traffic or the traffic packets in the transmission are equal to the bandwidth-delay product, that is, the cache space set in the network device is full of traffic cache, and the outflow and inflow of traffic in the cache space are equal.
  • the bandwidth allocation is maintained in a stable state to make the bandwidth resource allocation of each network device more reasonable.
  • the traffic in the data forwarding channel is respectively identified by the traffic identification and classification module to identify the type of traffic (traffic type), and the key value (Key) of each group of traffic is determined.
  • the occupied bandwidth (traffic bandwidth) of each group of traffic is counted by the traffic bandwidth statistics module, and the traffic type, key value, traffic bandwidth, etc. are sorted out to obtain a flow information table.
  • the cache calculation management module (device cache calculation management) is used to filter out the required data from the flow information table.
  • the cache calculation management module determines the bandwidth value allocated to the Cubic traffic at the current moment (time t) or the cache space threshold (preset scheduling cache threshold) of the network device cache BBR traffic based on the data, and determines the bandwidth value allocated to the scheduled BBR traffic based on the preset scheduling cache threshold.
  • a cache allocation method is provided. Compared with the different preemption capabilities of each congestion control algorithm for bandwidth resources in the prior art, bandwidth resources cannot be reasonably allocated according to demand, which reduces the fairness of bandwidth preemption of different congestion control algorithms.
  • the congestion control algorithm used by different network devices when sending traffic is identified, and the occupied bandwidth information of the traffic is counted; the cache allocation value of the network device is determined based on the congestion control algorithm, the occupied bandwidth information and the cache allocation rule; the traffic of the network device is allocated to the cache based on the cache allocation value.
  • the cache allocation value to be allocated to the network device is determined according to the congestion control algorithm, the occupied bandwidth information and the cache allocation rule, so as to accurately allocate the network traffic to the network device according to the cache allocation value. That is, in this application, the cache allocation value to be allocated to the network device is determined according to the congestion control algorithm, the occupied bandwidth information and the cache allocation rule, and the traffic of the network device is allocated to the cache according to the cache allocation value, so as to reasonably allocate bandwidth resources according to demand and improve the fairness of bandwidth preemption of different congestion control algorithms.
  • FIG. 4 is a flow chart of the second embodiment of the picture sharing method of the present application.
  • Step S01 obtaining the cache capacity of the traffic cache area in the network device and the egress bandwidth of each output port in the network device;
  • Step S02 determining the device type of the network device based on a preset classification cache threshold, the cache amount and the egress bandwidth;
  • Step S03 determining a cache allocation rule based on the device type.
  • the speed at which the congestion control algorithm occupies bandwidth to send traffic is greater than the transmission speed of the traffic, the traffic sent through the congestion control algorithm needs to be cached in the cache area of the network device to avoid packet loss.
  • different network devices use different congestion control algorithms. According to the characteristics of the congestion control algorithm occupying bandwidth, the cache space of the network bandwidth can be classified, and then the network devices can be classified to select the cache allocation rules that need to be used when determining the cache allocation value of the network device, so as to save data processing steps and improve the efficiency of bandwidth allocation.
  • the preset classification cache threshold is first determined according to the demand, and the cache capacity of the traffic cache area in the network device and the export bandwidth that can be occupied by the network device summary output port are obtained. The quotient of the cache capacity and the export bandwidth is calculated, and then the quotient is multiplied by the preset classification cache threshold to determine the device type of the network device. Finally, according to the device type, it is selected whether to use the scheduling cache rule in the cache allocation rule or to select the cache allocation value rule.
  • the device types include Cubic algorithm dominant devices and BBR algorithm dominant devices;
  • the step of determining the device type of the network device based on the preset classification cache threshold, the cache amount and the egress bandwidth includes:
  • Step S021 calculating the quotient of the cache amount and the egress bandwidth to obtain a relative cache value of the network device
  • Step S022 comparing the relative cache value with a preset classification cache threshold
  • Step S023 if the relative cache value is greater than the preset classification cache threshold, determining that the network device is the Cubic algorithm leading device;
  • Step S024 If the relative cache value is less than or equal to the preset classification cache threshold, it is determined that the device type of the network device is the BBR algorithm dominant device.
  • the Cubic algorithm can obtain more bandwidth resources by preempting the cache, and its performance in preempting bandwidth is significantly better than that of the BBR algorithm; however, when the cache space is small, since the bandwidth that the Cubic algorithm can preempt is smaller, and the BBR algorithm does not need to perceive available resources by preempting bandwidth, its performance is significantly better than that of the Cubic algorithm. Therefore, the congestion control algorithm that dominates the transmission traffic of the network device can be determined by the size of the cache space, and the congestion control algorithm that needs to be suppressed can be determined, that is, the bandwidth of the traffic needs to be allocated so that the traffic will not be congested in the cache space and cause packet loss.
  • step S20 of the present embodiment further includes:
  • Step S211 generating a real-time updated flow information table based on the congestion control algorithm and the occupied bandwidth information
  • Step S212 filters out the total port cache value, the egress bandwidth, the Cubic occupied bandwidth, the number of irrelevant flows, the number of Cubic flows, the number of other flows and the number of BBR flows from the flow information table;
  • Step S213 determining the cache allocation value of the network device based on the total port cache value, the egress bandwidth, the Cubic occupied bandwidth, the number of irrelevant flows, the number of Cubic flows, the number of remaining flows, the number of BBR flows and the cache allocation rule.
  • a flow information table can be generated based on the congestion control algorithm and the occupied bandwidth information, but the data required to determine the cache allocation value needs to be filtered out from the congestion control algorithm and the occupied bandwidth information.
  • a flow information table is generated based on the congestion control algorithm and the occupied bandwidth information, so that the cache calculation management module can quickly find the required data according to the sorted directory of the flow information table, thereby improving the efficiency of determining the cache allocation value.
  • the cache allocation rule when calculating the cache allocation value, it is necessary to first select the cache allocation rule required to determine the cache allocation value of the network device according to the device type, wherein the cache allocation rule includes at least a scheduling cache rule and a cache allocation value rule, and the scheduling cache rule is used to determine the cache threshold (preset scheduling threshold) that limits the BBR traffic in the cache space, and then control the sending rate, so as to avoid excessive traffic grabbing by network devices with BBR traffic as the device type, so that the Cubic traffic in the network device cannot be sent; or the cache allocation value rule is used to determine the bandwidth that can be used by the Cubic traffic during the sending process, so as to suppress the Cubic algorithm from occupying too much bandwidth, resulting in the network device's cache space being too full, causing packet loss. That is, in this embodiment, by controlling the preset scheduling threshold of the BBR traffic and the cache allocation value of the Cubic traffic, the bandwidth resources are reasonably allocated according to demand, and the fairness of bandwidth grabbing of different congestion control algorithms is improved.
  • the cache threshold pre
  • FIG6 is a flow chart of the third embodiment of the picture sharing method of the present application.
  • the cache allocation rule includes a scheduling cache rule
  • the congestion control algorithm includes a BBR algorithm.
  • step S20 further includes:
  • Step S1 if the device type of the network device is the BBR algorithm dominant device, then filter out from the occupied bandwidth information the total cache value of the port connected to the network device, the egress bandwidth, the BBR occupied bandwidth of the BBR traffic at the current moment, the number of other traffic items that use a congestion control algorithm other than the BBR algorithm to transmit traffic, and the number of BBR flows of the BBR traffic;
  • the BBR traffic is the traffic transmitted using the BBR algorithm
  • Step S2 determining a preset scheduling cache threshold of the network device based on the scheduling cache rule, the total port cache value, the egress bandwidth, the BBR occupied bandwidth, the number of remaining traffic items and the number of BBR flows;
  • the preset scheduling threshold belongs to the cache allocation value.
  • BBR traffic occupies most of the bandwidth. Since the BBR algorithm hardly occupies the cache, it is impossible to limit the rate of BBR traffic by displaying the BBR cache or increasing the Cubic cache. Therefore, it is necessary to delay the scheduling of traffic through network devices, that is, to make the BBR traffic wait in the cache space.
  • the queue in the scheduling cache space reaches a certain length (the preset scheduling cache threshold)
  • the BBR traffic is sent according to the queue scheduling bandwidth. Since the BBR algorithm is a congestion scheduling algorithm based on round-trip delay, the BBR traffic sending rate is actively reduced through delayed scheduling, thereby reducing the bandwidth occupied by BBR.
  • a scheduling cache rule is selected, wherein the scheduling cache rule may be a calculation formula for a preset scheduling cache threshold, and the preset scheduling cache threshold is calculated using the calculation formula for the preset scheduling threshold.
  • the calculation formula of the preset scheduling buffer threshold is:
  • Qb(t) is the preset scheduling cache threshold. If it is lower than this threshold, BBR traffic will not be scheduled. At this time, the newly entered BBR traffic will be added to the cache until the queue length exceeds the preset scheduling cache threshold.
  • Q is the cache amount
  • B is the egress bandwidth of the network device port
  • Bb(t) is the bandwidth occupied by BBR at time t
  • Nc(t) is the number of Cubic traffic items at time t (the number of TCP sessions using the Cubic algorithm), that is, the number of other traffic items
  • Nb(t) is the number of BBR traffic items at time t (the number of TCP sessions using the BBR algorithm).
  • the BBR algorithm will occupy more bandwidth in competition with the Cubic algorithm.
  • the above formula monitors the bandwidth occupied by BBR traffic in real time and sets the preset scheduling cache threshold of BBR traffic based on real-time bandwidth information. Increasing this threshold is equivalent to increasing the round-trip time (RTT) of BBR traffic.
  • RTT round-trip time
  • the sender of BBR traffic will reduce the sending traffic.
  • Qb(t) drops to 0.
  • the threshold value will also change dynamically, realizing a dynamic fairness of BBR and Cubic traffic bandwidth preemption.
  • the cache allocation rule further includes a cache allocation value rule
  • the congestion control algorithm further includes a Cubic algorithm.
  • step 20 in this embodiment further includes:
  • Step S3 if the device type of the network device is a Cubic algorithm dominant device, then filter out the total port cache value, the egress bandwidth, the Cubic occupied bandwidth occupied by the Cubic traffic at the current moment, the number of Cubic flows of the Cubic traffic at the current moment, and the number of irrelevant flows that use the Cubic algorithm to transmit traffic from the occupied bandwidth information;
  • the Cubic traffic is the traffic transmitted based on the Cubic algorithm
  • Step S4 determining the cache value of the Cubic traffic that needs to be allocated to the network device at the current moment based on the cache allocation value rule, the total port cache value, the egress bandwidth, the Cubic occupied bandwidth, the number of Cubic flows, and the number of irrelevant flows;
  • the cache value belongs to the cache allocation value.
  • the Cubic algorithm when the relative cache value is greater than the preset classification cache threshold, the Cubic algorithm will occupy more bandwidth by monitoring the bandwidth occupied by the Cubic traffic in real time and dynamically allocating it to the Cubic traffic cache space using the cache allocation value rule until the bandwidth occupied by the Cubic traffic is equal to the average loan. In this way, the dynamic cache allocation value is determined and the cache allocation is maintained in this stable state. If the traffic or bandwidth changes, the cache allocation value needs to be re-determined.
  • the cache allocation rule is determined to be a cache allocation value rule, wherein the cache allocation value rule may be a calculation formula for the cache allocation value, and the cache allocation value is calculated using the calculation formula for the cache allocation value.
  • the calculation formula of the cache allocation value may be:
  • Qc(t) is the cache value allocated to Cubic traffic at time t, It is a custom value and can be customized by the network administrator or user according to the actual situation.
  • Bc(t) is the total bandwidth occupied by all Cubic traffic at time t.
  • Nc(t) is the number of Cubic traffic at time t (the number of TCP sessions using the Cubic algorithm), and Nb(t) is the number of BBR traffic at time t (the number of TCP sessions using the BBR algorithm), that is, the number of irrelevant flows.
  • the packets of the Cubic traffic are discarded randomly by the low-priority queue or other discarding methods, while the BBR scheduling method is not affected and still uses the high-priority priority scheduling and the first-in-first-out principle of the same priority.
  • the present application also provides a cache allocation device.
  • the cache allocation device includes:
  • the information collection module 701 is used to identify the congestion control algorithm used by different network devices when sending traffic, and to collect statistics on the occupied bandwidth information of the traffic;
  • a determination module 702 configured to determine a cache allocation value of the network device based on the congestion control algorithm, the occupied bandwidth information and a cache allocation rule;
  • the allocation module 703 is used to allocate the traffic of the network device to the cache based on the cache allocation value.
  • the determination module 702 is also used to obtain the cache capacity of the traffic cache area in the network device and the export bandwidth of each output port in the network device; determine the device type of the network device based on a preset classification cache threshold, the cache capacity and the export bandwidth; and determine the cache allocation rules based on the device type.
  • the device types include Cubic dominant devices and BBR dominant devices;
  • the determination module 702 is also used to calculate the quotient of the cache amount and the export bandwidth to obtain the relative cache value of the network device; compare the relative cache value with the preset category cache threshold; if the relative cache value is greater than the preset category cache threshold, determine that the network device is the Cubic dominant device; if the relative cache value is less than or equal to the preset category cache threshold, determine that the network device is the BBR dominant device.
  • the cache allocation rule includes a scheduling cache rule, and the congestion control algorithm includes a BBR algorithm;
  • the determination module 702 is also used to filter out the total cache value of the port connected to the network device, the egress bandwidth, the BBR occupied bandwidth of the BBR traffic at the current moment, the number of remaining traffic items that use congestion control algorithms other than the BBR algorithm to transmit traffic, and the number of BBR flows of the BBR traffic from the occupied bandwidth information if the device type of the network device is the BBR algorithm-dominated device; wherein the BBR traffic is traffic transmitted using the BBR algorithm; based on the scheduling cache rule, the total cache value of the port, the egress bandwidth, the BBR occupied bandwidth, the number of remaining traffic items and the number of BBR flows, determine the preset scheduling cache threshold of the network device; wherein the preset scheduling threshold belongs to the cache allocation value.
  • the cache allocation rule includes a cache allocation value rule
  • the congestion control algorithm also includes a Cubic algorithm
  • the determination module 702 is also used to filter out the total port cache value, the egress bandwidth, the Cubic occupied bandwidth occupied by the Cubic traffic at the current moment, the number of Cubic flows of the Cubic traffic at the current moment, and the number of irrelevant flows that use the Cubic algorithm to transmit the traffic from the occupied bandwidth information if the device type of the network device is a Cubic algorithm-dominated device; wherein the Cubic traffic is the traffic transmitted based on the Cubic algorithm; based on the cache allocation value rule, the total port cache value, the egress bandwidth, the Cubic occupied bandwidth, the number of Cubic flows and the number of irrelevant flows, determine the cache value of the Cubic traffic that needs to be allocated to the network device at the current moment; wherein the cache value belongs to the cache allocation value.
  • the information collection module 701 is also used to extract the window trend characteristics of the traffic sent by each network device; compare the window trend characteristics with different types of preset theoretical window trends to determine the congestion control algorithm used to send the traffic.
  • the determination module 702 is also used to generate a real-time updated flow information table based on the congestion control algorithm and the occupied bandwidth information; filter out the port total cache value, the egress bandwidth, the Cubic occupied bandwidth, the number of irrelevant flows, the number of Cubic flows, the number of remaining flows and the number of BBR flows from the flow information table; determine the cache allocation value of the network device based on the port total cache value, the egress bandwidth, the Cubic occupied bandwidth, the number of irrelevant flows, the number of Cubic flows, the number of remaining flows, the number of BBR flows and the cache allocation rule.
  • An embodiment of the present application provides a storage medium, and the storage medium stores one or more programs, and the one or more programs can also be executed by one or more processors to implement the steps of any of the above-mentioned cache allocation methods.
  • the technical solution of the present application can be embodied in the form of a software product, which is stored in a storage medium (such as ROM/RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal device (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods of each embodiment of the present application.
  • a storage medium such as ROM/RAM, a magnetic disk, or an optical disk
  • a terminal device which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本申请公开了一种网络缓存分配方法、装置以及设备，涉及网络管理技术领域，该方法包括：识别网络设备发送流量时所采用的拥塞控制算法，并统计流量的占用带宽信息；基于拥塞控制算法、占用带宽信息和缓存分配规则确定网络设备的缓存分配值；基于缓存分配值将网络设备的流量分配缓存。

Description

缓存分配方法、装置以及设备
本申请要求于2023年8月16日申请的、申请号为202311037769.6的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及网络传输管理技术领域，尤其涉及一种缓存分配方法、装置以及设备。
 背景技术
随着网络设备的种类越来越多，为了使每种网络设备在使用时减少数据流的拥堵，常使用不同的拥塞控制算法从边缘网络设备中抢占网络带宽资源。
每台边缘网络设备都连接有多台网络设备，由于边缘网络设备可使用的带宽资源是有限的，且每种拥塞控制算法对带宽资源的抢占能力也是不同，所以会导致使用强抢占带宽资源能力强的拥塞控制算法的网络设备占用的带宽资源多，使用强抢占带宽资源能力弱的拥塞控制算法的网络设备占用的带宽资源少，导致带宽资源不能根据需求合理分配，降低了不同拥塞控制算法带宽抢占的公平性。
 技术问题
本申请的主要目的在于提供一种缓存分配方法、装置以及设备，旨在解决现有技术中带宽资源不能根据需求合理分配，降低了不同拥塞控制算法带宽抢占的公平性的技术问题。
 技术解决方案
为实现上述目的，本申请提供一种缓存分配方法，所述缓存分配方法，包括：
识别网络设备发送流量时所采用的拥塞控制算法，并统计所述流量的占用带宽信息；
基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值；
基于所述缓存分配值将所述网络设备的流量分配缓存。
在一实施例中，所述基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值的步骤之前，还包括：
获取所述网络设备中流量缓存区的缓存量，以及所述网络设备中每个输出端口的出口带宽；
基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型；
基于所述设备类型，确定缓存分配规则。
在一实施例中，所述设备类型包括Cubic算法主导设备与BBR算法主导设备；
所述基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型的步骤，包括：
对所述缓存量与所述出口带宽求商，得到所述网络设备的相对缓存值；
将所述相对缓存值与预设分类缓存门限进行比较；
若所述相对缓存值大于所述预设分类缓存门限，则确定所述网络设备为所述Cubic算法主导设备；
若所述相对缓存值小于或等于所述预设分类缓存门限，则确定所述网络设备为所述BBR算法主导设备。
在一实施例中，所述缓存分配规则包括调度缓存规则，所述拥塞控制算法包括BBR算法，所述基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值的步骤，包括：
若所述网络设备的设备类型为所述BBR算法主导设备，则从所述占用带宽信息中筛选出连接所述网络设备的端口总缓存值、出口带宽、当前时刻BBR流量的BBR占用带宽、采用除所述BBR算法之外的拥塞控制算法传输流量的其余流量条数，以及所述BBR流量的BBR流条数；
其中，所述BBR流量为采用所述BBR算法传输的流量；
基于所述调度缓存规则、所述端口总缓存值、所述出口带宽、所述BBR占用带宽、所述其余流量条数与所述BBR流条数，确定所述网络设备的预设调度缓存门限；
其中，所述预设调度门限属于所述缓存分配值。
在一实施例中，所述缓存分配规则还包括缓存分配值规则，所述拥塞控制算法还包括Cubic算法，所述基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值的步骤，包括：
若所述网络设备的设备类型为Cubic算法主导设备，则从所述占用带宽信息中筛选出所述端口总缓存值、所述出口带宽、当前时刻Cubic流量占用的Cubic占用带宽、当前时刻所述Cubic流量的Cubic流条数，以及采用除所述Cubic算法传输流量的无关流条数；
其中，所述Cubic流量为基于所述Cubic算法传输的流量；
基于所述缓存分配值规则、所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述Cubic流条数以及所述无关流条数，确定当前时刻需要向所述网络设备分配的Cubic流量的缓存分配值；
其中，所述缓存值属于所述缓存分配值。
在一实施例中，所述识别网络设备发送流量时所采用的拥塞控制算法的步骤，包括：
提取各网络设备发送的流量的窗口趋势特征；
将所述窗口趋势特征与不同类型的预设理论窗口趋势对比，确定发送所述流量采用的拥塞控制算法。
在一实施例中，所述基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值的步骤，还包括：
基于所述拥塞控制算法与所述占用带宽信息，生成实时更新的流信息表；
从所述流信息表中筛选出所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述无关流条数、所述Cubic流条数、所述其余流量条数与所述BBR流条数。
基于所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述无关流条数、所述Cubic流条数、所述其余流量条数、所述BBR流条数与缓存分配规则，确定所述网络设备的缓存分配值。
此外，为实现上述目的，本申请还提供一种缓存分配装置，缓存分配装置包括：
信息采集模块，用于识别不同网络设备发送流量时所采用的拥塞控制算法，并统计所述流量的占用带宽信息；
确定模块，用于基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值；
分配模块，用于基于所述缓存分配值将所述网络设备的流量分配缓存。
在一实施例中，所述确定模块，还用于获取所述网络设备中流量缓存区的缓存量，以及所述网络设备中每个输出端口的出口带宽；基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型；基于所述设备类型，确定缓存分配规则。
此外，为实现上述目的，本申请还提出一种缓存分配设备，所述设备包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的缓存分配程序，所述缓存分配程序配置为实现如上文所述的缓存分配方法的步骤。
此外，为实现上述目的，本申请还提出一种存储介质，所述存储介质上存储有缓存分配程序，所述缓存分配程序被处理器执行时实现如上文所述的缓存分配方法的步骤。
 有益效果
本申请提供一种缓存分配方法、装置以及设备，与现有技术中每种拥塞控制算法对带宽资源的抢占能力不同，带宽资源不能根据需求合理分配，降低了不同拥塞控制算法带宽抢占的公平性相比，在本申请中，识别不同网络设备发送流量时所采用的拥塞控制算法，并统计所述流量的占用带宽信息；基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值；基于所述缓存分配值将所述网络设备的流量分配缓存。在本申请中，在识别不同网络设备发送流量时所采用的拥塞控制算法，并统计流量的占用带宽信息后，根据拥塞控制算法、占用带宽信息和缓存分配规则确定需要向网络设备分配的缓存分配值，以根据缓存分配值精准将网络设备分配网络的流量分配缓存，即在本申请中，根据拥塞控制算法、占用带宽信息与缓存分配规则确定需要向网络设备分配的缓存分配值，按照缓存分配值将网络设备的流量分配缓存，以将带宽资源根据需求合理分配，提高不同拥塞控制算法带宽抢占的公平性。
 附图说明
此处的附图被并入说明书中并构成本说明书的一部分，示出了符合本申请的实施例，并与说明书一起用于解释本申请的原理。
为了更清楚地说明本申请实施例或现有技术中的技术方案，下面将对实施例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地，对于本领域普通技术人员而言，在不付出创造性劳动的前提下，还可以根据这些附图获得其它的附图。
图1为本申请实施例方案涉及的硬件运行环境的设备机构示意图；
图2为本申请缓存分配方法第一实施例的流程示意图；
图3为本申请缓存分配方法的实际工作的流程示意图；
图4为本申请缓存分配方法第二实施例的流程示意图；
图5为本申请缓存分配方法中的网络设备的设备分类流程示意图；
图6为本申请缓存分配方法第三实施例的流程示意图；
图7为本申请缓存分配装置的结构配置示意图。
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。
 本发明的实施方式
应当理解，此处所描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。
参照图1，图1为本申请实施例方案涉及的硬件运行环境的缓存分配设备结构示意图。
如图1所示，该缓存分配设备可以包括：处理器1001，例如中央处理器（Central Processing Unit，CPU），通信总线1002、用户接口1003，网络接口1004，存储器1005。其中，通信总线1002用于实现这些组件之间的连接通信。用户接口1003可以包括显示屏（Display）、输入单元比如键盘（Keyboard），用户接口1003还可以包括标准的有线接口、无线接口。网络接口1004可以包括标准的有线接口、无线接口（如无线保真（WIreless-FIdelity，WI-FI）接口）。存储器1005可以是高速的随机存取存储器（Random Access Memory，RAM）存储器，也可以是稳定的非易失性存储器（Non-Volatile Memory，NVM），例如磁盘存储器。存储器1005还可以是独立于前述处理器1001的存储装置。
本领域技术人员可以理解，图1中示出的结构并不构成对缓存分配设备的限定，可以包括比图示更多或更少的部件，或者组合某些部件，或者不同的部件布置。
如图1所示，作为一种存储介质的存储器1005中可以包括操作系统、数据存储模块、网络通信模块、用户接口模块以及缓存分配程序。
在图1所示的缓存分配设备中，网络接口1004主要用于与其他设备进行数据通信；用户接口1003主要用于与用户进行数据交互；本申请缓存分配设备中的处理器1001、存储器1005可以设置在缓存分配设备中，所述缓存分配设备通过处理器1001调用存储器1005中存储的缓存分配程序，并执行本申请实施例提供的缓存分配方法。
本申请实施例提供了一种缓存分配方法，参照图2，图2为本申请一种缓存分配方法第一实施例的流程示意图。
需要说明的是，本实施例的执行主体可以是所述缓存分配设备，所述缓存分配设备可以是个人电脑、智能手机、平板电脑等电子设备的边缘网络设备，还可以是其他可实现相同或相似功能的其他设备，本实施例对此不加以限制，在本实施例及下述各实施例中，以缓存分配设备为例对本申请缓存分配方法进行说明。
需要说明的是，网络带宽分配设备中还包括流量识别分类模块（图中未示出）、流量带宽统计模块（图中未示出）、设备缓存计算管理模块（图中未示出），还可以包括流信息表（参考图3）。
其中，流量识别分类模块用于识别网络设备及发送流量时所采用的拥塞控制算法，并根据拥塞控制算法对流量进行分类记录，还可以根据流量五元组信息通过哈希算法，确定每组流量的键值；流量带宽统计模块可以实时监控各种流量的占用带宽信息；流信息表用于记录流量识别分类模块等得到流量信息与流量带宽统计模块得到的占用带宽信息，并根据流量信息与占用带宽信息实时更新；设备缓存计算管理模块用于根据流量信息表执行缓存计算和分配，并下发到数据层面执行。
本实施例中，所述缓存分配方法包括：
步骤S10，识别网络设备发送流量时所采用的拥塞控制算法，并统计所述流量的占用带宽信息；
其中，网络设备可以是电脑、服务器、手机或平板等可以利用网络进行数据传输的电子设备，具体不做限定。
其中，拥塞控制算法可以是一种用来调整传输控制协议(TCP)连续单次发送数据分组数量的算法，可以通过增减单次发送数据分组数量逐步调整网络带宽，使之逼近当前网络的承载量。
其中，拥塞控制算法可以包括Cubic（立方的）算法、BBR（Bottleneck Bandwidth and Round-trip time，并将带宽和往返时间）算法等，具体不做限定。
需要说明的是，由于每种拥塞控制算法对带宽的抢占方式不同，且对带宽的抢占强度也不同，为了避免一些拥塞控制算法过度抢占带宽，使其余的拥塞控制算法无法传输流量，所以需要识别网络设备传输流量采用的拥塞控制算法，并统计将每个拥塞控制流量的占用带宽信息。其中，拥塞控制流量可看做是采用哪种拥塞控制算法传输流量，就将该流量看做是该拥塞控制算法。
例如，采用Cubic算法传输的流量可以看做是Cubic流量；采用BBR算法传输的流量可以看做是BBR流量。
在具体实现中，在接收到所有与自身连接的网络设备的流量后，利用流量识别分类模块确定网络设备发送流量的窗口趋势特征，根据窗口趋势特征，识别网络设备传输流量时所采用的拥塞控制算法，并利用流量带宽统计模块统计，各网络设备占用的流量带宽，也即，每个拥塞控制流量的占用带宽信息。
进一步地，为了能够提高识别网络设备发送流量时所采用的拥塞控制算法的效率，本实施例中步骤S10，还包括：
步骤S11，提取各网络设备发送的流量的窗口趋势特征；
步骤S12，将所述窗口趋势特征与不同类型的预设理论窗口趋势对比，确定发送所述流量采用的拥塞控制算法。
需要说明的是，窗口可以看做是流量发生拥塞的拥塞窗口；窗口趋势可以看做是流量传输时拥塞窗口处的拥堵趋势；预设理论窗口趋势可以是通过实验数据得到的拥塞窗口流量的拥堵趋势。
在具体实现中，先利用流量识别分类模块收集实验数据，并通过实验数据整理出Cubic算法的Cubic窗口趋势与BBR算法的BBR窗口趋势，提取网络设备发送流量的窗口趋势特征，通过窗口趋势特征分别与预设理论窗口趋势对比，判定发送流量所采用的拥堵控制算法，也即，通过实验得到数据确定各拥塞控制算法的拥塞窗口处的拥塞趋势，并利用该拥塞趋势判断网络设备发送流量所采用的拥塞控制算法，避免了复杂计算过程，提高了识别发送流的拥塞控制算法的效率。
步骤S20，基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值；
需要说明的是，确定每组流量对应的拥塞控制算法，并根据每组流行的占用带宽信息与带宽分配规则，确定每个网络设备可以使用的最大带宽，并根据最大带宽，确定每个网络设备的缓存分配值，以使用每种拥塞控制算法的网络设备，都能将得到可以满足自身需求的带宽资源，以避免某一种拥塞控制算法过多占用带宽资源，造成使用其他拥塞控制算法的网络设备无法使用带宽或可使用带宽太少无法正常发送流量的情况，提高了不同拥塞控制算法对带宽抢占的公平性。
在具体实现中，利用缓存计算管理模块确定当前时刻各拥塞控制算法所对应的占用带宽，并从占用带宽信息分析出各网络设备端口可以缓存流量的总缓存大小、各网络设备端口的出口带宽、在当前时刻每种拥塞控制算法占用的总带宽与每种流量的流条数，利用缓存分配规则在缓存计算管理模块中，确定各网络设备的缓存分配值，以根据缓存分配值合理分配带宽资源。
其中，网络设备端口的出口带宽可以是网络设备发送流量时可以占用的最多的带宽，若超出该带宽，可能会出现端口拥挤，因为拥挤出现丢包的情况。
步骤S30，基于所述缓存分配值将所述网络设备的流量分配缓存。
需要说明的是，缓存分配值可以是零值，也即，不需要向该网络设备分配带宽，或调整该网络设备的流量缓存区域。
需要说明的是，根据缓存分配值向网络设备分配网络带宽，可以实时在分配过程中实时检测网络设备的带宽占用情况，直至网络设备的占据带宽等于流量平均带宽或者传输途中的流量包等于带宽时延积，也即，网路设备中设置的缓存空间被流量缓存满，且缓存空间内流量的流出与流入持平，在占据带宽等于流量平均带宽或缓存空间内流量的流出与流入持平时，维持该带宽分配的稳定状态，以使每个网络设备的带宽资源分配更加合理。
在具体实现中，参考图3，数据转发通道中的流量分别通过流量识别分类模块识别流量的类型（流量类型），并确定出每组流量的键值（Key），通过流量带宽统计模块统计每组流量的占用带宽（流量带宽），并将流量类型、键值、流量带宽等整理后得到流信息表，利用缓存计算管理净模块（设备缓存计算管理）从流信息表中筛选出需要用到的数据，缓存计算管理模块根据该数据确定出在当前时刻（t时刻），分配给Cubic流量的带宽值，或网络设备缓存BBR流量的缓存空间门限（预设调度缓存门限），并根据预设调度缓存门限确定分配给调度BBR流量的带宽值。
在本实施例中提供一种缓存分配方法，与现有技术中每种拥塞控制算法对带宽资源的抢占能力不同，带宽资源不能根据需求合理分配，降低了不同拥塞控制算法带宽抢占的公平性相比，在本申请中，识别不同网络设备发送流量时所采用的拥塞控制算法，并统计所述流量的占用带宽信息；基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值；基于所述缓存分配值将所述网络设备的流量分配缓存。在本申请中，在识别不同网络设备发送流量时所采用的拥塞控制算法，并统计流量的占用带宽信息后，根据拥塞控制算法、占用带宽信息和缓存分配规则确定需要向网络设备分配的缓存分配值，以根据缓存分配值精准将网络设备分配网络的流量分配缓存，即在本申请中，根据拥塞控制算法、占用带宽信息与缓存分配规则确定需要向网络设备分配的缓存分配值，按照缓存分配值将网络设备的流量分配缓存，以将带宽资源根据需求合理分配，提高不同拥塞控制算法带宽抢占的公平性。
参考图4，图4为本申请图片分享方法第二实施例的流程示意图。
基于上述实施例，在本实施例中，所述步骤20之前，包括：
步骤S01，获取所述网络设备中流量缓存区的缓存量，以及所述网络设备中每个输出端口的出口带宽；
步骤S02，基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型；
步骤S03，基于所述设备类型，确定缓存分配规则。
需要说明的是，由于拥塞控制算法占用带宽发送流量的速度大于流量的传输速度，所以通过拥塞控制算法发送的流量需要缓存在网络设备的缓存区以避免丢包，但是不同的网络设备采用的拥塞控制算法也不同，根据拥塞控制算法抢占带宽的特性，可以将网络带宽的缓存空间进行分类，进而对网络设备进行分类，以选定确定该网络设备的缓存分配值时需要使用缓存分配规则，以节省数据处理步骤，提高带宽分配的效率。
在具体实现中，先根据需求确定预设分类缓存门限，并获取网络设备中流量缓存区的缓存量与网络设备汇总输出端口可以占用的出口带宽，通过求缓存量与出口带宽的商，再与预设分类缓存门限，确定网络设备的设备类型，最后根据设备类型选择是使用缓存分配规则中的调度缓存规则，还是选择缓存分配值规则。
进一步地，所述设备类型包括Cubic算法主导设备与BBR算法主导设备；
所述基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型的步骤，包括：
步骤S021，对所述缓存量与所述出口带宽求商，得到所述网络设备的相对缓存值；
步骤S022，将所述相对缓存值与预设分类缓存门限进行比较；
步骤S023，若所述相对缓存值大于所述预设分类缓存门限，则确定所述网络设备为所述Cubic算法主导设备；
步骤S024，若所述相对缓存值小于或等于所述预设分类缓存门限，则确定所述网络设备的设备类型为所述BBR算法主导设备。
需要说明的是，由于在网络设备的缓存空间比较大时，Cubic算法通过抢占缓存可以获得更多的带宽资源，其抢占带宽的性能明显优于BBR算法；但是，在缓存空间较小时，由于Cubic算法可抢占的带宽较小，而BBR算法无需通过抢占带宽感知可用资源，其性能明显优于Cubic算法，所以通过缓存空间的大小即可判断出主导所述网络设备传输流量的拥塞控制算法，并确定需要抑制的拥塞控制算法，也即，需要分配流量的带宽，以使流量不会在缓存空间拥堵，造成丢包。
在实际实行中，参照图5，利用缓存计算管理模块，计算出换存量Q与出口带宽B的商，得到网络设备的相对缓存值C，也即，C=Q/B，再将相对缓存值与预设分类缓存门限M进行对比，如果M大于C，则确定网络设备为Cubic主导设备；如果M不大于C，也即，M小于或等于C，则确定主导网络设备传输流量的算法为BBR算法，也即，网络设备为BBR主导设备，以根据这两种设备类型确定带宽分配规律的选择。
进一步地，为了提高缓存分配值的准确率，所述本实施例的步骤S20，还包括：
步骤S211，基于所述拥塞控制算法与所述占用带宽信息，生成实时更新的流信息表；
步骤S212从所述流信息表中筛选出所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述无关流条数、所述Cubic流条数、所述其余流量条数与所述BBR流条数；
步骤S213，基于所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述无关流条数、所述Cubic流条数、所述其余流量条数、所述BBR流条数与缓存分配规则，确定所述网络设备的缓存分配值。
需要说明的是，可以根据拥塞控制算法与占用带宽信息生成流信息表，但是需要从拥塞控制算法与占用带宽信息中筛选出确定缓存分配值需要的数据，在本实施例中，根据拥塞控制算法与占用带宽信息生成流信息表，使缓存计算管理模块可以根据流信息表的整理目录快速找到需要的数据，以提高确定缓存分配值的效率。
在具体实现中，计算缓存分配值时需要先根据设备类型选择确定网络设备的缓存分配值时需要用到的缓存分配规则，其中，缓存分配规则至少包括调度缓存规则与缓存分配值规则，利用调度缓存规则确定限制BBR流量在缓存空间的缓存门限（预设调度门限）进而控制发送的速率，以避免设备类型为BBR流量的网络设备过度的抢占流量，使该网络设备中的Cubic流量无法发送；或利用缓存分配值规则确定Cubic流量在发送过程中可以使用的带宽，进而抑制Cubic算法占用过多的带宽，导致网络设备的缓存空间过满，造成丢包。也即，在本实施例中，通过控制BBR流量的预设调度门限与Cubic流量的缓存分配值，使带宽资源根据需求合理分配，提高不同拥塞控制算法带宽抢占的公平性。
参考图6，图6为本申请图片分享方法第三实施例的流程示意图。
基于上述实施例，在本实施例中，所述缓存分配规则包括调度缓存规则，所述拥塞控制算法包括BBR算法，在缓存分配规则为调度缓存规则，所述拥塞控制算法为BBR算法时，所述步骤S20，还包括：
步骤S1，若所述网络设备的设备类型为所述BBR算法主导设备，则从所述占用带宽信息中筛选出连接所述网络设备的端口总缓存值、出口带宽、当前时刻BBR流量的BBR占用带宽、采用除所述BBR算法之外的拥塞控制算法传输流量的其余流量条数，以及所述BBR流量的BBR流条数；
其中，所述BBR流量为采用所述BBR算法传输的流量；
步骤S2，基于所述调度缓存规则、所述端口总缓存值、所述出口带宽、所述BBR占用带宽、所述其余流量条数与所述BBR流条数，确定所述网络设备的预设调度缓存门限；
其中，所述预设调度门限属于所述缓存分配值。
需要说明的是，当相对缓存值小于货等于预设分类缓存门限时，BBR流量占用大部分带宽，由于BBR算法几乎不占用缓存，所以无法通过显示BBR缓存或增加Cubic缓存的方式限制BBR流量的速率，所以需要通过网络设备延迟对流量的调度，也即，使BBR流量现在缓存空间内等待排队，在调度缓存空间内的队列达到一定长度（预设调度缓存门限）时，再根据队列调度带宽发送BBR流量，由于BBR算法是基于往返时延的拥塞调度算法，所以通过延迟调度，使BBR流量发送速率主动降低，从而降低BBR占据的带宽。
在具体实现中，在确定设备类型为BBR主导设备时，选择调度缓存规则，其中调度缓存规则可以是预设调度缓存门限的计算公式，利预设调度门限的计算公式，计算出预设调度缓存门限。
其中，预设调度缓存门限的计算公式为：
其中，Qb(t)为预设调度缓存门限，低于该门限条件下不对BBR流量进行调度，此时新进入设备的BBR流量会陆续加入缓存，直到队列长度高于预设调度缓存门限 为自定义系数, Q为缓存量，B为网络设备端口的出口带宽，Bb(t)为t时刻BBR占用带宽，Nc(t)为t时刻Cubic流量条数(采用Cubic算法的TCP会话数量)，也即，其余流量条数，Nb(t)为t时刻BBR流量条数(采用BBR算法的TCP会话数量)。
需要说明的是，在缓存空间相对较小的网络设备中，BBR算法在与Cubic算法竞争中会占据更多的带宽，上述公式通过实时监控BBR流量所占据的带宽并根据实时带宽信息设定BBR流量的预设调度缓存门限。增加该门限之后等同于增加了BBR流量的往返时间(RTT)，根据BBR的拥塞管理机制可知增加RTT时BBR流量发送端会降低发送流量，当流量降低到平均带宽时Qb(t)降为0, 当流量发生变化时该门限值也会动态发生变化，实现BBR和Cubic流量带宽抢占的一种动态公平性。
在一实施例中，所述缓存分配规则还包括缓存分配值规则，所述拥塞控制算法还包括Cubic算法，在缓存分配规则为缓存分配值规则时，本实施例中步骤20，还包括：
步骤S3，若所述网络设备的设备类型为Cubic算法主导设备，则从所述占用带宽信息中筛选出所述端口总缓存值、所述出口带宽、当前时刻Cubic流量占用的Cubic占用带宽、当前时刻所述Cubic流量的Cubic流条数，以及采用除所述Cubic算法传输流量的无关流条数；
其中，所述Cubic流量为基于所述Cubic算法传输的流量；
步骤S4，基于所述缓存分配值规则、所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述Cubic流条数以及所述无关流条数，确定当前时刻需要向所述网络设备分配的Cubic流量的缓存值；
其中，所述缓存值属于所述缓存分配值。
需要说明的是，当相对缓存值大于预设分类缓存门限时，Cubic算法会占据更多的带宽，通过实时监控Cubic流量所占用的带宽，并利用缓存分配值规则动态分配给Cubic流量缓存空间，直到Cubic流量占据的带宽等于平均贷款后，以确定动态的缓存分配值，并维持缓存分配该稳定状态，若流量或带宽发生变化，则需要重新确定缓存分配值。
在具体实现中，网络设备的设备类型为Cubic主导设备，则确定缓存分配规则为缓存分配值规则，其中，缓存分配值规则可以是缓存分配值的计算公式，利用缓存分配值的计算公式计算缓存分配值。
其中，缓存分配值的计算公式可以为：
其中，Qc(t)为t时刻分配给Cubic流量的缓存值, 为自定义值，可以由网络管理员或用户根据实际情况自定义配置。Bc(t)为t时刻所有Cubic流量占据的总带宽。Nc(t)为t时刻Cubic流量条数(采用Cubic算法的TCP会话数量)，Nb(t)为t时刻BBR流量条数(采用BBR算法的TCP会话数量)，也即，无关流条数。
需要说明的是，当Cubic流量占满缓存空间时开始，以低优先级队列随机丢弃或其他丢弃方式丢弃Cubic流量的报文，而BBR调度方式不受影响，依然采用高优先级优先调度，同优先级先进先出原则。
本申请还提供一种缓存分配装置，参考图7，缓存分配装置包括：
信息采集模块701，用于识别不同网络设备发送流量时所采用的拥塞控制算法，并统计所述流量的占用带宽信息；
确定模块702，用于基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值；
分配模块703，用于基于所述缓存分配值将所述网络设备的流量分配缓存。
在一实施例中，所述确定模块702，还用于获取所述网络设备中流量缓存区的缓存量，以及所述网络设备中每个输出端口的出口带宽；基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型；基于所述设备类型，确定缓存分配规则。
在一实施例中，所述设备类型包括Cubic主导设备和BBR主导设备；
所述确定模块702，还用于对所述缓存量与所述出口带宽求商，得到所述网络设备的相对缓存值；将所述相对缓存值与预设分类缓存门限进行比较；若所述相对缓存值大于所述预设分类缓存门限，则确定所述网络设备为所述Cubic主导设备；若所述相对缓存值小于或等于所述预设分类缓存门限，则确定所述网络设备为所述BBR主导设备。
在一实施例中，所述缓存分配规则包括调度缓存规则，所述拥塞控制算法包括BBR算法；
所述确定模块702，还用于若所述网络设备的设备类型为所述BBR算法主导设备，则从所述占用带宽信息中筛选出连接所述网络设备的端口总缓存值、出口带宽、当前时刻BBR流量的BBR占用带宽、采用除所述BBR算法之外的拥塞控制算法传输流量的其余流量条数，以及所述BBR流量的BBR流条数；其中，所述BBR流量为采用所述BBR算法传输的流量；基于所述调度缓存规则、所述端口总缓存值、所述出口带宽、所述BBR占用带宽、所述其余流量条数与所述BBR流条数，确定所述网络设备的预设调度缓存门限；其中，所述预设调度门限属于所述缓存分配值。
在一实施例中，所述缓存分配规则包括缓存分配值规则，所述拥塞控制算法还包括Cubic算法；
所述确定模块702，还用于若所述网络设备的设备类型为Cubic算法主导设备，则从所述占用带宽信息中筛选出所述端口总缓存值、所述出口带宽、当前时刻Cubic流量占用的Cubic占用带宽、当前时刻所述Cubic流量的Cubic流条数，以及采用除所述Cubic算法传输流量的无关流条数；其中，所述Cubic流量为基于所述Cubic算法传输的流量；基于所述缓存分配值规则、所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述Cubic流条数以及所述无关流条数，确定当前时刻需要向所述网络设备分配的Cubic流量的缓存值；其中，所述缓存值属于所述缓存分配值。
在一实施例中，所述信息采集模块701，还用于提取各网络设备发送的流量的窗口趋势特征；将所述窗口趋势特征与不同类型的预设理论窗口趋势对比，确定发送所述流量采用的拥塞控制算法。
在一实施例中，所述确定模块702，还用于基于所述拥塞控制算法与所述占用带宽信息，生成实时更新的流信息表；从所述流信息表中筛选出所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述无关流条数、所述Cubic流条数、所述其余流量条数与所述BBR流条数；基于所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述无关流条数、所述Cubic流条数、所述其余流量条数、所述BBR流条数与缓存分配规则，确定所述网络设备的缓存分配值。
本申请缓存分配装置的具体实施方式与上述缓存分配方法各实施例基本相同，在此不再赘述。
本申请实施例提供了一种存储介质，且存储介质存储有一个或者一个以上程序，一个或者一个以上程序还可被一个或者一个以上的处理器执行以用于实现上述任一项的缓存分配方法的步骤。
本申请存储介质具体实施方式与上述缓存分配方法各实施例基本相同，在此不再赘述。
需要说明的是，在本文中，术语“包括”、“包含”或者其任何其它变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、方法、物品或者装置不仅包括那些要素，而且还包括没有明确列出的其它要素，或者是还包括为这种过程、方法、物品或者装置所固有的要素。在没有更多限制的情况下，由语句“包括一个等”限定的要素，并不排除在包括该要素的过程、方法、物品或者装置中还存在另外的相同要素。
上述本申请实施例序号仅仅为了描述，不代表实施例的优劣。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到上述实施例方法可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件，但很多情况下前者是更佳的实施方式。基于这样的理解，本申请的技术方案本质上或者说对现有技术做出贡献的部分可以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质（如ROM/RAM、磁碟、光盘）中，包括若干指令用以使得一台终端设备（可以是手机，计算机，服务器，空调器，或者网络设备等）执行本申请各个实施例的方法。
以上仅为本申请的可选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效机构或等效流程变换，或直接或间接运用在其它相关的技术领域，均同理包括在本申请的专利保护范围内。

Claims (10)

  1. 一种缓存分配方法，其中，所述缓存分配方法，包括：
    识别网络设备发送流量时所采用的拥塞控制算法，并统计所述流量的占用带宽信息；
    基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值；
    基于所述缓存分配值将所述网络设备的流量分配缓存。
  2. 如权利要求1所述的缓存分配方法，其中，所述基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值的步骤之前，还包括：
    获取所述网络设备中流量缓存区的缓存量，以及所述网络设备中每个输出端口的出口带宽；
    基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型；
    基于所述设备类型，确定缓存分配规则。
  3. 如权利要求2所述的缓存分配方法，其中，所述设备类型包括Cubic算法主导设备与BBR算法主导设备；
    所述基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型的步骤，包括：
    对所述缓存量与所述出口带宽求商，得到所述网络设备的相对缓存值；
    将所述相对缓存值与预设分类缓存门限进行比较；
    若所述相对缓存值大于所述预设分类缓存门限，则确定所述网络设备为所述Cubic算法主导设备；
    若所述相对缓存值小于或等于所述预设分类缓存门限，则确定所述网络设备为所述BBR算法主导设备。
  4. 如权利要求3所述的缓存分配方法，其中，所述缓存分配规则包括调度缓存规则，所述拥塞控制算法包括BBR算法，所述基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值的步骤，包括：
    若所述网络设备的设备类型为所述BBR算法主导设备，则从所述占用带宽信息中筛选出连接所述网络设备的端口总缓存值、出口带宽、当前时刻BBR流量的BBR占用带宽、采用除所述BBR算法之外的拥塞控制算法传输流量的其余流量条数，以及所述BBR流量的BBR流条数；
    其中，所述BBR流量为采用所述BBR算法传输的流量；
    基于所述调度缓存规则、所述端口总缓存值、所述出口带宽、所述BBR占用带宽、所述其余流量条数与所述BBR流条数，确定所述网络设备的预设调度缓存门限；
    其中，所述预设调度门限属于所述缓存分配值。
  5. 如权利要求4所述的缓存分配方法，其中，所述缓存分配规则还包括缓存分配值规则，所述拥塞控制算法还包括Cubic算法，所述基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值的步骤，包括：
    若所述网络设备的设备类型为Cubic算法主导设备，则从所述占用带宽信息中筛选出所述端口总缓存值、所述出口带宽、当前时刻Cubic流量占用的Cubic占用带宽、当前时刻所述Cubic流量的Cubic流条数，以及采用除所述Cubic算法传输流量的无关流条数；
    其中，所述Cubic流量为基于所述Cubic算法传输的流量；
    基于所述缓存分配值规则、所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述Cubic流条数以及所述无关流条数，确定当前时刻需要向所述网络设备分配的Cubic流量的缓存值；
    其中，所述缓存值属于所述缓存分配值。
  6. 如权利要求1至5任一项所述的缓存分配方法，其中，所述识别网络设备发送流量时所采用的拥塞控制算法的步骤，包括：
    提取各网络设备发送的流量的窗口趋势特征；
    将所述窗口趋势特征与不同类型的预设理论窗口趋势对比，确定发送所述流量采用的拥塞控制算法。
  7. 如权利要求5所述的缓存分配方法，其中，所述基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值的步骤，还包括：
    基于所述拥塞控制算法与所述占用带宽信息，生成实时更新的流信息表；
    从所述流信息表中筛选出所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述无关流条数、所述Cubic流条数、所述其余流量条数与所述BBR流条数；
    基于所述端口总缓存值、所述出口带宽、所述Cubic占用带宽、所述无关流条数、所述Cubic流条数、所述其余流量条数、所述BBR流条数与缓存分配规则，确定所述网络设备的缓存分配值。
  8. 一种缓存分配装置，其中，所述缓存分配装置包括：
    信息采集模块，用于识别不同网络设备发送流量时所采用的拥塞控制算法，并统计所述流量的占用带宽信息；
    确定模块，用于基于所述拥塞控制算法、所述占用带宽信息和缓存分配规则确定所述网络设备的缓存分配值；
    分配模块，用于基于所述缓存分配值向所述网络设备分配缓存。
  9. 如权利要求8所述的缓存分配装置，其中，所述确定模块，还用于获取所述网络设备中流量缓存区的缓存量，以及所述网络设备中每个输出端口的出口带宽；基于预设分类缓存门限、所述缓存量与所述出口带宽，确定所述网络设备的设备类型；基于所述设备类型，确定缓存分配规则。
  10. 一种缓存分配设备，其中，所述缓存分配设备包括：存储器、处理器及存储在所述存储器上并可在所述处理器上运行的缓存分配程序，所述缓存分配程序配置为实现如权利要求1至7中任一项所述的缓存分配方法的步骤。
PCT/CN2023/131355 2023-08-16 2023-11-13 缓存分配方法、装置以及设备 WO2025035627A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202311037769.6 2023-08-16
CN202311037769.6A CN117041151A (zh) 2023-08-16 2023-08-16 缓存分配方法、装置以及设备

Publications (1)

Publication Number Publication Date
WO2025035627A1 true WO2025035627A1 (zh) 2025-02-20

Family

ID=88637006

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/131355 WO2025035627A1 (zh) 2023-08-16 2023-11-13 缓存分配方法、装置以及设备

Country Status (2)

Country Link
CN (1) CN117041151A (zh)
WO (1) WO2025035627A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117041151A (zh) * 2023-08-16 2023-11-10 鹏城实验室 缓存分配方法、装置以及设备

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018001222A1 (zh) * 2016-06-27 2018-01-04 中兴通讯股份有限公司 数据传输缓存队列分配方法及装置和计算机存储介质
CN114726790A (zh) * 2021-01-06 2022-07-08 郑州芯兰德网络科技有限公司 一种基于传输内容大小的拥塞控制选择方法及系统
CN115865827A (zh) * 2023-02-09 2023-03-28 合肥综合性国家科学中心人工智能研究院(安徽省人工智能实验室) 一种基于事件驱动的跨层端网协同拥塞控制方法
CN116155825A (zh) * 2023-04-04 2023-05-23 山东大学 一种bbr拥塞控制算法数据重传的优化方法
CN117041151A (zh) * 2023-08-16 2023-11-10 鹏城实验室 缓存分配方法、装置以及设备

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2018001222A1 (zh) * 2016-06-27 2018-01-04 中兴通讯股份有限公司 数据传输缓存队列分配方法及装置和计算机存储介质
CN114726790A (zh) * 2021-01-06 2022-07-08 郑州芯兰德网络科技有限公司 一种基于传输内容大小的拥塞控制选择方法及系统
CN115865827A (zh) * 2023-02-09 2023-03-28 合肥综合性国家科学中心人工智能研究院(安徽省人工智能实验室) 一种基于事件驱动的跨层端网协同拥塞控制方法
CN116155825A (zh) * 2023-04-04 2023-05-23 山东大学 一种bbr拥塞控制算法数据重传的优化方法
CN117041151A (zh) * 2023-08-16 2023-11-10 鹏城实验室 缓存分配方法、装置以及设备

Also Published As

Publication number Publication date
CN117041151A (zh) 2023-11-10

Similar Documents

Publication Publication Date Title
CN106789660B (zh) 软件定义网络中QoS可感知的流量管理方法
KR100429904B1 (ko) 차등화된 ＱｏＳ 서비스를 제공하는 라우터 및 그것의고속 ＩＰ 패킷 분류 방법
EP2671354B1 (en) System to share network bandwidth among competing applications
CN103229466B (zh) 一种数据包传输的方法及装置
US8891543B1 (en) Method and system for processing packets in a network device
JP2003152792A (ja) パケット転送装置、方法およびプログラム
CN103442076B (zh) 一种云存储系统的可用性保障方法
CN106789721B (zh) 一种基于令牌桶的智能qos方法及系统
CN109905329B (zh) 一种虚拟化环境下任务类型感知的流队列自适应管理方法
US8203956B1 (en) Method and apparatus providing a precedence drop quality of service (PDQoS)
CN111108728A (zh) 对报文进行处理方法及装置
WO2025035627A1 (zh) 缓存分配方法、装置以及设备
CN114079638A (zh) 多协议混合网络的数据传输方法、装置和存储介质
CN115037694A (zh) 一种数据传输方法、装置、电子设备和存储介质
US20210176182A1 (en) Method and Apparatus for Determining Packet DeQueue Rate
KR101737516B1 (ko) 공평한 대역 할당 기반 패킷 스케줄링 방법 및 장치
CN111131061B (zh) 一种数据传输方法及网络设备
CN113765812A (zh) 一种标记报文的方法和装置
CN111756586A (zh) 一种数据中心网络中基于优先级队列的公平带宽分配方法、交换机及可读存储介质
US20070058649A1 (en) Packet queuing system and method
CN112055382A (zh) 一种基于精细化区分的业务接入方法
KR20120055947A (ko) 가입자 인지 플로우별 ＱｏＳ 제공 방법 및 장치
US11182205B2 (en) Multi-processor queuing model
CN113765796A (zh) 流量转发控制方法及装置
US8612647B2 (en) Priority aware queue

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23949045

Country of ref document: EP

Kind code of ref document: A1