WO2018072690A1 - 用于测量网络路径的方法、装置和系统 - Google Patents

用于测量网络路径的方法、装置和系统 Download PDF

Info

Publication number
WO2018072690A1
WO2018072690A1 PCT/CN2017/106557 CN2017106557W WO2018072690A1 WO 2018072690 A1 WO2018072690 A1 WO 2018072690A1 CN 2017106557 W CN2017106557 W CN 2017106557W WO 2018072690 A1 WO2018072690 A1 WO 2018072690A1
Authority
WO
WIPO (PCT)
Prior art keywords
switching node
available bandwidth
switching
path
aggregate
Prior art date
Application number
PCT/CN2017/106557
Other languages
English (en)
French (fr)
Inventor
胡农达
向海洲
涂伯颜
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Priority to EP17862716.2A priority Critical patent/EP3515018B1/en
Publication of WO2018072690A1 publication Critical patent/WO2018072690A1/zh
Priority to US16/385,954 priority patent/US11082358B2/en

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/70—Admission control; Resource allocation
    • H04L47/76—Admission control; Resource allocation using dynamic resource allocation, e.g. in-call renegotiation requested by the user or requested by the network in response to changing network conditions
    • H04L47/762—Admission control; Resource allocation using dynamic resource allocation, e.g. in-call renegotiation requested by the user or requested by the network in response to changing network conditions triggered by the network
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0876—Network utilisation, e.g. volume of load or congestion level
    • H04L43/0882—Utilisation of link capacity
    • 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
    • H04L47/11—Identifying congestion
    • 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
    • H04L47/12—Avoiding congestion; Recovering from congestion
    • 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
    • H04L47/82—Miscellaneous aspects
    • H04L47/822—Collecting or measuring resource availability data
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00—Packet switching elements
    • H04L49/15—Interconnection of switching modules
    • H04L49/1515—Non-blocking multistage, e.g. Clos
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L49/00—Packet switching elements
    • H04L49/50—Overload detection or protection within a single switching element
    • H04L49/505—Corrective measures
    • H04L49/506—Backpressure

Definitions

  • the present invention relates to the field of network technologies, and in particular, to a method, apparatus, and system for measuring a network path.
  • FIGS. 1 and 2 show schematic diagrams of two different stages of CLOS switching networks: a Spine-Leaf switching network and a Fat-Tree switching network.
  • the circles in FIGS. 1 and 2 represent computing nodes, the rectangles represent switching nodes, and the straight lines represent links between nodes.
  • the step-by-step back pressure method is a commonly used congestion measurement method.
  • the switching nodes transmit the congestion information in a reverse direction along each path, thereby determining the congestion status of each path between the pair of edge switching nodes through the congestion information.
  • the congestion information can have multiple representation methods, such as the available bandwidth of the path.
  • the network path measurement method based on the step-by-step back pressure method is not accurate enough.
  • the present invention provides a method, apparatus and system for measuring a network path that can improve the efficiency of measuring a network path.
  • a first aspect a method for measuring a network path, comprising: obtaining, by a first switching node in a CLOS switching network, a first aggregate available bandwidth of a path of the first switching node to a second switching node,
  • the first switching node and the second switching node are peer switching nodes in the CLOS switching network;
  • the first switching node acquires a path of the first target port of the third switching node to the first switching node a first available bandwidth
  • the third switching node is a next-level switching node connected to the first switching node;
  • the first switching node acquires a second path of the second switching node to a fourth switching node
  • the available bandwidth, the fourth switching node is a next-level switching node connected to the second switching node;
  • the first switching node determines a first target port of the third switching node to the fourth switching node a second aggregate available bandwidth of the path, the second aggregate available bandwidth being the smallest of the first aggregate available bandwidth, the first available bandwidth, and the
  • the first switching node in the CLOS switching network can calculate the aggregate available bandwidth of the path between the first target port of the at least one next-level switching node connected to the next lower-level switching node connected to the next lower-level switching node in the above manner, so as to be the next level.
  • the switching node dynamically schedules traffic between ports according to the aggregate available bandwidth of each port to implement load balancing. This method reduces the complexity of measuring the network path and improves the efficiency of measuring the network path.
  • the first switching node in the CLOS switching network acquires the first aggregate available bandwidth of the path from the first switching node to the second switching node, including: the first switching node respectively From multiple The fifth switching node receives the first information, where the first information indicates an aggregate available bandwidth of the first path corresponding to each of the plurality of fifth switching nodes, where the multiple fifth switching nodes Is a higher-level switching node connected to the first switching node, where the first path is a path through the first switching node to the second switching node of each fifth switching node; The first switching node determines, according to the first information, a sum of aggregate available bandwidths of the first path corresponding to the multiple fifth switching nodes as the first aggregate available bandwidth.
  • the first switching node and the second switching node are next-level switching nodes connected to multiple sixth switching nodes in the CLOS switching network, and the CLOS switching network Obtaining, by the first switching node, the first aggregate available bandwidth of the path of the first switching node to the second switching node, including: the first switching node acquiring the first switching node to the multiple sixth exchanges An uplink available bandwidth of a path of each sixth switching node in the node; the first switching node receives second information from each of the sixth switching nodes, and the second information is used to indicate each of the sixth a downlink available bandwidth of the switching node to the second switching node; the first switching node determines that the smaller available bandwidth of the uplink available bandwidth and the downlink available bandwidth is available for each sixth switching node Bandwidth; determining a sum of available bandwidths corresponding to the plurality of sixth switching nodes as the first aggregate available bandwidth.
  • the first switching node acquires a second available bandwidth of the path from the second switching node to the fourth switching node, where: the first switching node receives the second switching node Sending third information, where the third information is used to indicate that the available bandwidth of the path from the second switching node to the fourth switching node is the second available bandwidth.
  • the method further includes: the first switching node sends fourth information to the third switching node, where the fourth information is used to indicate that the first target port of the third switching node is The available bandwidth between the fourth switching nodes is the second aggregate available bandwidth.
  • a method for measuring a network path including: obtaining, by a third switching node in a CLOS switching network, a first available path of a path from a first target port of the third switching node to a first switching node Bandwidth, the third switching node is a next-level switching node connected to the first switching node; the third switching node acquires a third aggregate available bandwidth of a path of the first switching node to the fourth switching node The fourth switching node and the third switching node are peer switching nodes; the third switching node determines a second path of the first switching port of the third switching node to the fourth switching node The available bandwidth is aggregated, and the second aggregate available bandwidth is a smaller available bandwidth of the first available bandwidth and the third aggregate available bandwidth.
  • the third switching node in the CLOS switching network can calculate the aggregate available bandwidth between its first target port and the fourth switching node in the above manner. Therefore, the third switching node dynamically schedules traffic between the ports according to the aggregate available bandwidth corresponding to each port to implement load balancing. Improve the efficiency of measuring network paths.
  • the third aggregate available bandwidth is a smaller available bandwidth of the first aggregate available bandwidth and the second available bandwidth, where the first aggregate available bandwidth is the first switch node to a second available bandwidth of the path of the second switching node, the second available bandwidth is a second available bandwidth of the path of the second switching node to the fourth switching node, and the second switching node is The upper switching node connected by the four switching nodes.
  • the third switching node acquires a third aggregate available bandwidth of the path of the first switching node to the fourth switching node, including: the third switching node from the first switching The node receives the first indication information, where the first indication information is used to indicate that the aggregate available bandwidth of the path of the first switching node to the fourth switching node is the third aggregate available bandwidth.
  • the third switching node acquires the first switching node to the fourth switching node.
  • the third aggregated available bandwidth of the path includes: the third switching node receives the second indication information sent by the first switching node, and the second indication information is used to indicate the first switching node to the second exchange a first aggregated available bandwidth of the path of the node; the third switching node receives the third indication information sent by the second switching node, where the third indication information is used to indicate the second switching node to the fourth a second available bandwidth of the path of the switching node; the third switching node determines the third aggregate available bandwidth, the third aggregate available bandwidth is a smaller available bandwidth of the first aggregated available bandwidth and the second available bandwidth .
  • the method further includes: determining, by the third switching node, a fourth aggregate available bandwidth of the third switching node to the fourth switching node, where the fourth aggregate available bandwidth is the The sum of the aggregated available bandwidths of the paths of the ports of the three switching nodes to the paths of the fourth switching nodes, respectively.
  • an apparatus comprising means for performing the method of the first aspect.
  • the apparatus comprising means for performing the method of the first aspect.
  • an apparatus comprising means for performing the method of the second aspect.
  • the apparatus comprising means for performing the method of the second aspect.
  • a switching network system comprising the apparatus of the above third aspect and the apparatus of the above fourth aspect.
  • an apparatus comprising: a memory for storing a program; a transceiver for communicating with other devices; and a processor for executing a program in the memory, when the program is executed, The processor is operative to perform the method of the first aspect when the program is executed.
  • an apparatus comprising: a memory for storing a program; a transceiver for communicating with other devices; and a processor for executing a program in the memory, when the program is executed, The processor is operative to perform the method of the second aspect when the program is executed.
  • a switching network system comprising the apparatus of the sixth aspect, and the apparatus of the seventh aspect.
  • a system chip comprising means for performing the method of the first aspect.
  • a system chip comprising means for performing the method of the second aspect.
  • a switching network system comprising the system chip of the above ninth aspect, and the system chip of the above tenth aspect.
  • FIG. 1 is a schematic structural diagram of a Spine-Leaf switching network according to an embodiment of the present invention
  • FIG. 2 is a schematic structural diagram of a Fat-Tree switching network according to an embodiment of the present invention.
  • FIG. 3 is a schematic diagram of a method for measuring a network path based on back pressure according to an embodiment of the present invention
  • FIG. 4 is a schematic diagram of a method for measuring a network path based on back pressure according to another embodiment of the present invention.
  • FIG. 5 is a schematic structural diagram of a multi-stage CLOS switching network according to another embodiment of the present invention.
  • FIG. 6 is a schematic block diagram of a method for measuring a network path according to another embodiment of the present invention.
  • FIG. 7 is a schematic block diagram of a method for measuring a network path according to still another embodiment of the present invention.
  • FIG. 8 is a schematic topological diagram of a CLOS switching sub-network according to still another embodiment of the present invention.
  • FIG. 9 is a schematic diagram of a method for measuring a network path according to still another embodiment of the present invention.
  • FIG. 10 is a schematic diagram of a method for measuring a network path according to still another embodiment of the present invention.
  • FIG. 11 is a schematic diagram of a method for measuring a network path according to still another embodiment of the present invention.
  • FIG. 12 is a schematic diagram of a method for measuring a network path according to still another embodiment of the present invention.
  • FIG. 13 is a schematic structural diagram of a multi-stage CLOS network according to still another embodiment of the present invention.
  • FIG. 14 is a schematic structural diagram of a multi-stage CLOS network according to still another embodiment of the present invention.
  • FIG. 15 is a schematic illustration of an apparatus in accordance with an embodiment of the present invention.
  • Figure 16 is a schematic illustration of an apparatus in accordance with another embodiment of the present invention.
  • FIG. 17 is a schematic illustration of an apparatus in accordance with another embodiment of the present invention.
  • Figure 18 is a schematic illustration of an apparatus in accordance with another embodiment of the present invention.
  • the bandwidth in the embodiment of the present invention may be represented by an absolute bandwidth value, or may be represented by a quantized value.
  • the bandwidth may be expressed in bits/second (bits), or may be represented by a bandwidth quantization value.
  • bits/second bits
  • the manner in which the bandwidth is expressed in the embodiment of the present invention is not limited.
  • measuring the network path may generally refer to measuring the shortest network path between the switching nodes.
  • the CLOS switching network is a special multistage Interconnection Networks.
  • the CLOS switching network may include the Spine-Leaf switching network shown in FIG. 1 and/or the Fat-Tree switching network shown in FIG. 2.
  • the multi-level CLOS switching network in the present invention may be constructed by: taking two sets of switching nodes, and each switching node in the first group is connected to each switching node in the second group, Construct a secondary CLOS switching network.
  • the secondary CLOS switching network is regarded as a virtual switching node, and a plurality of virtual switching nodes are formed into a group, and each virtual switching node is connected to each switching node of the third group of switching nodes, which can form a third level. CLOS switching network.
  • each virtual switching node may be referred to as a pod.
  • an i-level CLOS switching network is regarded as a virtual switching node, and a plurality of virtual switching nodes are formed into a group, and each virtual switching node is connected to each of the i+1th switching nodes. It can form an i+1 level CLOS switching network.
  • the switching node in the first group of switching nodes is called an edge switching node
  • the switching node in the i+1th group switching node is called a core switching node or a root switching node
  • the switching nodes in the second group to the i-th switching node are called intermediate switching nodes.
  • the edge switching node may be referred to as a TOR switching node (represented by T ij )
  • the core switching node is referred to as a Core switching node (represented by C ij )
  • the intermediate switching node is called the Agg switching node (represented by A ij ).
  • edge switching nodes are connected to compute nodes (eg, servers).
  • Figure 3 shows how the amount of congestion of multiple paths of edge switching nodes T 1.1 to T 2.1 is implemented based on step-by-step backpressure.
  • the congestion status is measured by the quantized value of the available bandwidth. The larger the value, the larger the available bandwidth and the lower the congestion.
  • Each link uses "(link uplink available bandwidth, link downlink available bandwidth)" to indicate the available bandwidth in both directions of the link.
  • the measurement process is described.
  • the switching node T2.1 passes through the upper port P3 (it is required that the P* in the embodiment of the present invention is used to indicate the * port of the corresponding switching node, for example, P3.
  • the indication information is stored, and the available bandwidth information is further fed back to the upper hop switching nodes C 1.1 and C 1.2 .
  • the available bandwidth information fed back at this time needs to be updated according to the available bandwidth information of the link between A 2.1 and the previous hop switching node, and the update method is: taking the received available bandwidth value and the minimum available bandwidth of the link. .
  • the available bandwidth of the bottleneck link on the path is taken as the available bandwidth of the path.
  • This process can continue until the available bandwidth information is fed back to the edge switching node T 1.1 .
  • path information may be recorded in the indication information, and when the indication information is finally fed back to the edge node, the edge node may distinguish different paths.
  • the switching node T 1.1 will receive the available bandwidth information of the four different paths back to the switching node T 2.1 .
  • Figure 4 shows these four different paths and their available bandwidth.
  • the available bandwidth is represented by the number inside the circle. Based on this method, the available bandwidth of all the shortest paths between any two switching nodes can be measured.
  • the above-mentioned step-by-step backpressure based network path measurement method has the following disadvantages: First, since the number of paths between the two edge switching nodes may be very large, transmitting congestion information may occupy a large amount of link bandwidth. For example, for a Fat-Tree network constructed by isomorphic switching nodes of m ports, the number of different paths between a pair of edge switching nodes can be as high as m 2 /4. Second, for multiple paths between a pair of edge-switched nodes, it is impossible to distinguish whether they share the available bandwidth. For example, as shown in FIG. 4, the available bandwidths of path 3 and path 4 from the upper port P4 of the switching node T 1.1 are 3 and 1, respectively, and the sum of the available bandwidths is 4.
  • the available bandwidth of the uplink of the shared link ⁇ T 2.1 , A 2.2 > of path 3 and path 4 is 3, so path 3 and path 4 share the available bandwidth. That is, the aggregate available bandwidth of the two paths is smaller than the sum of the available bandwidths of the two paths.
  • the available bandwidths of path 1 and path 2 from the upper port P3 of the switching node T 1.1 are 2 and 2, respectively, and the sum of the available bandwidths is 4. They share the available bandwidth of the physical link ⁇ T 1.1 , A 1.1 > 4, so path 1 and path 2 do not share the available bandwidth. That is, the aggregate available bandwidth of the two paths is equal to the sum of their available bandwidth.
  • the embodiment of the present invention provides a method, device and system for measuring a network path.
  • the program is based on the following core ideas:
  • the switching node (for example, T 1.1 in Figure 4) only needs to know how to allocate traffic on its outgoing ports, how much traffic should be allocated should not depend on how much bandwidth is available for the link connected to its outgoing port. It should depend on how much aggregate available bandwidth is available from each of its outgoing ports to the destination switching node (eg, T 1.1 to T 2.1 in Figure 4).
  • a set of paths from an outbound port of the source switching node to the destination switching node is regarded as an aggregation path, and the available bandwidth of the aggregation path may be referred to as an aggregate available bandwidth.
  • the aggregate available bandwidth of an outgoing port of the switching node to another switching node is often not equal to the sum of the available bandwidth of each path from the outgoing port to another switching node, because the available bandwidth may be shared among the multiple paths.
  • the available bandwidth of port P3 to T 2.1 of T 1.1 in Figure 4 is not necessarily equal to the path "T 1.1 -> A 1.1 -> C 1.1 -> A 2.1 -> T 2.1 " and path "T 1.1 -> A 1.1 ->C 1.2 ->A 2.1 ->T 2.1 "The sum of the available bandwidth, as these two paths may share the available bandwidth of link L ⁇ T 1.1 , A 1.1 > and link L ⁇ A 2.1 , T 2.1 > .
  • a correct method for calculating the available bandwidth of an aggregated path is to divide the aggregated path into a combination of series-parallel path segments, sum the available bandwidths of the parallel path segments, and obtain the available path segments for each path segment in series. The minimum bandwidth.
  • the aggregation path is divided into a combination of series-parallel path segments, and the calculation of the aggregate available bandwidth of the aggregation path is converted into the calculation of the aggregate available bandwidth of each path segment connected in parallel and the The aggregation of path segments can be a combination of available bandwidth calculations.
  • the calculation method for the aggregate available bandwidth of each path segment connected in parallel is to obtain the sum of the (aggregated) available bandwidths of the path segments
  • the calculation method for the aggregate available bandwidth of each path segment in the series is to obtain the path segments (aggregation) The minimum value of the available bandwidth.
  • the calculation of the aggregate available bandwidth for each path segment connected in parallel may have multiple definitions, at least two.
  • the first definition may be the summation of the available bandwidths of the parallel path segments as described above.
  • the second definition may be to average the available bandwidth of each path segment in parallel (ie, the sum of the available bandwidth of each path segment in parallel divided by the number of paths in parallel).
  • the first definition is adopted in the embodiment of the present invention.
  • the method for measuring the network path in the embodiment of the present invention may also be implemented, and details are not described herein again. It will be obvious to those skilled in the art that various modifications and changes can be made in the above-described manner of the available bandwidth of the aggregation, and such modifications or variations are also within the scope of the embodiments of the present invention.
  • FIG. 5 is a schematic diagram showing a path between a pair of edge nodes of a CLOS switching network according to an embodiment of the present invention, which shows a feature of hierarchically aggregating paths between switching nodes of a multi-stage CLOS switching network from the inside to the outside. As shown in FIG.
  • the paths between the two switching nodes in the CLOS switching network are hierarchically aggregated from the inside to the outside. This simplifies the measurement of the available bandwidth of the aggregate of the path. Therefore, the aggregation path between the source edge switching node and the destination edge switching node of the multi-level CLOS switching network can be divided into such a combination of serial-parallel path ends, and has a nesting characteristic, and a recursive method can be used to solve the bandwidth of the aggregate available path. .
  • the congestion state of the inner path can be measured first (for example, the congestion state is characterized by the available bandwidth), and then aggregated, and then extended and aggregated layer by layer to obtain the nodes of the same-level switching network in the CLOS switching network.
  • the aggregated available bandwidth information for the inter-path can obtain the aggregate available bandwidth of the CLOS switching network based on the recursive method.
  • FIG. 6 shows a schematic block diagram of a method for measuring a network path according to an embodiment of the present invention.
  • the method 600 can be performed by a switching node in a CLOS switching network. As shown in FIG. 6, the method 600 includes:
  • the first switching node in the CLOS switching network acquires a first aggregate available bandwidth of the path of the first switching node to the second switching node, where the first switching node and the second switching node exchange the CLOS A peer exchange node in the network.
  • the first switching node may be an intermediate switching node in the CLOS switching network.
  • the first switching node can be A 1.1 in FIG.
  • the first aggregate available bandwidth may be an aggregate available bandwidth of a set of paths between the first switching node and the second switching node.
  • the first aggregate available bandwidth is an aggregate available bandwidth of all paths between the first switching node and the second switching node.
  • the first aggregate available bandwidth is the actual available bandwidth of the aggregate path between the first switching node and the second switching node.
  • the first switching node acquires a first available bandwidth of a path of a first target port of the third switching node to the first switching node, where the third switching node is connected to the first switching node.
  • the next level of switching nodes is a first available bandwidth of a path of a first target port of the third switching node to the first switching node, where the third switching node is connected to the first switching node.
  • the third switching node is a next-level switching node connected to the first switching node.
  • the first switching node may be A 1.1 in FIG. 4
  • the third switching node may be 1.1 T in FIG. 4
  • the first target port may be a port P3 1.1 T in FIG. 4.
  • the path of the first target port of the third switching node to the first switching node may be an uplink of the third switching node to the first switching node.
  • the first switching node acquires a second available bandwidth of a path of the second switching node to a fourth switching node, where the fourth switching node is a next-level switching node that is connected to the second switching node.
  • the fourth switching node is a next-level switching node connected to the second switching node.
  • the second switching node may be a in FIG. 4 A 2.1, may be a fourth switching node in FIG. 4 T 2.1.
  • the path from the second switching node to the fourth switching node may be a downlink between the second switching node and the fourth switching node.
  • the first switching node determines a second aggregate available bandwidth of a path of the first target port of the third switching node to the fourth switching node, where the second aggregate available bandwidth is available for the first aggregation.
  • the bandwidth, the first available bandwidth, and the smallest available bandwidth of the second available bandwidth are the bandwidth, the first available bandwidth, and the smallest available bandwidth of the second available bandwidth.
  • the second aggregate available bandwidth may be an aggregate available bandwidth of the path of the first target port to the fourth switching node of the third switching node.
  • a switching node in a CLOS switching network can obtain an aggregate available bandwidth between each outbound port of a plurality of outbound ports to another switching node, so as to facilitate switching.
  • the node dynamically schedules traffic between ports to perform load balancing.
  • the first switching node in the CLOS switching network may calculate, by using the foregoing manner, the aggregation of the path between the first target port of the at least one lower-level switching node connected to the next lower-level switching node. Bandwidth, so that the next-level switching node dynamically schedules traffic between ports according to the aggregate available bandwidth of each port to implement load balancing. This method reduces the complexity of measuring the network path and improves the efficiency of measuring the network path.
  • the measurement of the network path may be performed in a periodic or event triggered manner.
  • the first switching node may receive information sent by the upper-level switching node to indicate that the first switching node is connected to the second switching node from the egress port connected to the upper-level node. Aggregate available bandwidth. The first switching node may determine the first aggregate available bandwidth of the path of the first switching node and the second switching node according to the sum of the aggregate available bandwidths of the multiple outbound ports.
  • the first switching node in the CLOS switching network acquires the first aggregate available bandwidth of the path of the first switching node to the second switching node, including: the first switching node separately from multiple fifth switching nodes Receiving, by the first information, an aggregate available bandwidth of the first path corresponding to each of the plurality of fifth switching nodes, where the multiple fifth switching nodes are a first-level switching node connected to the first switching node, where the first path is a path through the first switching node to the second switching node of each fifth switching node; the first switching The node determines, according to the first information, a sum of aggregate available bandwidths of the first path corresponding to the multiple fifth switching nodes as the first aggregate available bandwidth.
  • the first path may be a path or a set of paths of the first switching node to the second switching node.
  • the number of first paths depends on the number of paths from the first switching node to the second switching node passing through each fifth switching node.
  • the first path is the path between the first target port of the first switching node to the second switching node.
  • the first target port of a switching node may be an egress port through which the path of the first switching node to the fifth switching node passes.
  • the fifth switching node may be a root switching node in the CLOS switching network, or may be an intermediate switching node in the CLOS switching network.
  • the first path corresponding to each fifth switching node may include multiple paths.
  • the first path corresponding to each fifth switching node may be one path.
  • the first aggregated bandwidth between the first switching node and the second switching node may be determined by the following method:
  • the first switching node and the second switching node are next-level switching nodes connected to a plurality of sixth switching nodes in the CLOS switching network, and the first switching node in the CLOS switching network acquires the a first aggregated available bandwidth of the path from the first switching node to the second switching node, comprising: the first switching node acquiring the first switching node to each of the plurality of sixth switching nodes Uplink available bandwidth of the path; the first switching node receives second information from each of the sixth switching nodes, and the second information is used to indicate each of the sixth switching nodes to the second switching node a downlink available bandwidth; the first switching node determines that the smaller available bandwidth of the uplink available bandwidth and the downlink available bandwidth is the available bandwidth corresponding to each of the sixth switching nodes; The sum of the available bandwidths corresponding to the switching nodes is determined as the first aggregate available bandwidth.
  • the first switching node acquires the second available bandwidth of the path from the second switching node to the fourth switching node, where: the first switching node receives the second The third information sent by the switching node, where the third information is used to indicate that the available bandwidth of the path from the second switching node to the fourth switching node is the second available bandwidth.
  • the first switching node may send a message to the third switching node, indicating the first target of the third switching node. Aggregate available bandwidth between ports to other switching nodes.
  • the method 600 further includes: the first switching node sending fourth information to the third switching node, where the fourth information is used to indicate the first target port to the fourth exchange of the third switching node
  • the available bandwidth between the nodes is the second aggregate available bandwidth.
  • the method 600 in conjunction with FIG. 6, describes a method for calculating the aggregated available bandwidth of a path between ports of a next-level switching node by a higher-level switching node.
  • FIG. 7 shows a schematic block diagram of a method for measuring a network path according to another embodiment of the present invention.
  • a method 700 for measuring a network path in accordance with another embodiment of the present invention will now be described with reference to FIG.
  • the method 700 can be performed by a switching node in a CLOS switching network.
  • the method 700 describes a method for calculating the aggregated available bandwidth of a path between ports of a switching interface of the primary switching node by the primary switching node.
  • Method 700 includes:
  • the third switching node in the CLOS switching network acquires a first available bandwidth of a path of the first target port of the third switching node to the first switching node, where the third switching node is the first switching node.
  • the third switching node may be an intermediate switching node or an edge switching node in the CLOS switching network.
  • the third switching node may be T 1.1 in FIG. 4, and the first switching node may be A 1.1 in FIG.
  • the third switching node acquires a third aggregate available bandwidth of the path of the first switching node to the fourth switching node, where the fourth switching node and the third switching node are peer switching nodes.
  • the third switching node determines a second aggregate available bandwidth of a path from the first target port of the third switching node to the fourth switching node, where the second aggregate available bandwidth is the first available bandwidth. And a smaller available bandwidth of the third aggregate available bandwidth.
  • the third switching node in the CLOS switching network may calculate the aggregate available bandwidth between the first target port and the fourth switching node in the foregoing manner. Therefore, the third switching node dynamically schedules traffic between the ports according to the aggregate available bandwidth corresponding to each port to implement load balancing. Improve the efficiency of measuring network paths.
  • the third aggregate available bandwidth is a smaller available bandwidth of a first aggregate available bandwidth and a second available bandwidth, where the first aggregate available bandwidth is the first switch node to a second available bandwidth of the path of the second switching node, the second available bandwidth is a second available bandwidth of the path of the second switching node to the fourth switching node, and the second switching node is The upper switching node connected by the four switching nodes.
  • the third switching node acquires a third aggregate available bandwidth of the path of the first switching node to the fourth switching node, including: the third switching node from the first switching
  • the node receives the first indication information, where the first indication information is used to indicate that the aggregate available bandwidth of the path of the first switching node to the fourth switching node is the third aggregate available bandwidth.
  • the third switching node acquires a third aggregate available bandwidth of the path of the first switching node to the fourth switching node, where: the third switching node receives the first exchange a second indication information sent by the node, where the second indication information is used to indicate a first aggregate available bandwidth of the path of the first switching node to the second switching node; and the third switching node receives the second switching node a third indication information, where the third indication information is used to indicate a second available bandwidth of the path of the second switching node to the fourth switching node; the third switching node determines that the third aggregation is available Bandwidth, the third aggregate available bandwidth is a smaller available bandwidth of the first aggregate available bandwidth and the second available bandwidth.
  • the third switching node may further determine The aggregate available bandwidth of all paths from the three switching nodes to the fourth switching node.
  • the method further includes: the third switching node determines a fourth aggregate available bandwidth of the third switching node to the fourth switching node, and the fourth aggregate available bandwidth is the third exchange The sum of the aggregated available bandwidth of the path of each port of the node to the path of the fourth switching node.
  • the third switching node may indicate the aggregate available bandwidth of the third switching node to the fourth switching node to the next-level switching node, so that the next-level switching node continues to calculate the aggregate available bandwidth between the next-level switching nodes.
  • the method 700 may further include: the third switching node sends fourth indication information to the seventh switching node, where the fourth indication information is used to indicate between the third switching node and the fourth switching node.
  • the available bandwidth is the fourth aggregate available bandwidth
  • the seventh switching node is the next-level switching node of the third switching node.
  • FIG. 8 to FIG. 12 are merely for facilitating the understanding of the embodiments of the present invention, and the embodiments of the present invention are not limited to the specific numerical values or specific examples illustrated.
  • a person skilled in the art will be able to make various modifications and changes in accordance with the examples of FIG. 8 to FIG. 12, and such modifications and variations are also within the scope of the embodiments of the present invention.
  • the method for measuring the network path in the embodiment of the present invention can be divided into two types: one is the measurement of the aggregate available bandwidth of the path between the same-level switching nodes with the distance of 2; One is the measurement of the aggregate available bandwidth of a path between peers of the same level with a distance of 2k (k>1).
  • the distance between nodes refers to the number of links that pass from one switching node to another.
  • the above two types of methods can be performed in parallel.
  • FIG. 8 is taken as an example to describe how to perform measurement of aggregate available bandwidth of a path between peers of the same level with a distance of 2.
  • Figure 8 is a sub-picture of the Fat-Tree switching network topology shown in Figure 2, in which the distance between each switching node A *.1 is two.
  • 9 to 11 respectively show a method of measuring the aggregate available bandwidth of the path between the peer switching nodes in FIG.
  • the embodiments of the present invention respectively propose three methods for measuring the aggregate available bandwidth of a path between peer nodes of the same level.
  • each link is represented by "(link uplink available bandwidth, link downlink available bandwidth)" in the upper and lower directions.
  • the switching node C 1.* can obtain the available bandwidth information of each link by measuring the bandwidth occupation of the transmission/reception direction of each port. According to the available bandwidth information of each link, the available bandwidth of the path between the two downlink switching nodes can be calculated by taking the minimum value of the available bandwidth of the link connecting the two downlink switching nodes.
  • the available bandwidth of the path of port P3 of 2.1 is 2.
  • the switching nodes C 1.1 and C 1.2 can respectively calculate the path information table between the switching nodes A 1.* connected thereto, as shown in Table 1 and Table 2 in FIG.
  • the switching node C 1.* calculates the path information table between the switching nodes A *.1 , the corresponding path information (for example, the path information is equivalent to the first information in the method 600) is sent to The source switching node of the path. Therefore, each switching node A *.1 can obtain its available bandwidth information to other peer switching nodes. At the same time, because there may be multiple parallel paths between the two switching nodes, the switching node can also calculate the available bandwidth of the aggregation path to other switching nodes. The available bandwidth of the aggregation path is equal to the available bandwidth of all parallel paths between the two switching nodes. Sum. These switching nodes can store these path information in their path information tables.
  • the path information tables of A 1.1 , A 2.1, and A 3.1 are shown in Table 3, Table 4, and Table 5, respectively.
  • Table 4 there is a path from the ports P3 and P4 of the switching node A 2.1 to the switching node A 3.1 , and their available bandwidths are 2 and 3 respectively, so from the switching node A 2.1 to the switching node A 3.1
  • the aggregate available bandwidth is 5.
  • a switching node (which may be referred to as a "source switching node” for convenience of distinction) may send a downlink of a link connected to its upper port to other peer switching nodes to which the switching node is connected.
  • Available bandwidth information when other switching nodes at the same level receive the available bandwidth information, other switching nodes at the same level can calculate the path information of the source switching node to the downlink available bandwidth information.
  • the calculation method is: taking the minimum available bandwidth of the link connected to the port receiving the available bandwidth information and the received available bandwidth value. For example, as shown in FIG.
  • a 3.1 sends the downlink available bandwidth (value 6) of the link L ⁇ C 1.2 , A 2.1 > to the nodes A 1.1 and A 2.1 through the upper port P4; when the A 1.1 receives the available bandwidth information After that, compare it with the uplink available bandwidth (value 3) of link L ⁇ C 1.2 , A 1.1 > and take the minimum value (value 3), and obtain the available bandwidth of the path from port P4 to A 2.1 of A 1.1 . 3.
  • each switching node A *.1 can obtain the available bandwidth information of the path to other peer switching nodes.
  • the switching node can also calculate the aggregate available bandwidth of the aggregation path to other switching nodes.
  • the aggregate available bandwidth of the aggregation path is equal to all parallel paths between the two switching nodes.
  • These switching nodes store these path information in their path information tables.
  • the path information tables of A 1.1 , A 2.1, and A 3.1 are shown in Table 1, Table 2, and Table 3, respectively. They are identical to Table 3, Table 4 and Table 5 in Figure 9.
  • a switching node (which may be referred to as a "source switching node" for convenience of distinction) transmits downlink available links of links connected to the upper port to other peer switching nodes interconnected by the uplink switching node.
  • Bandwidth information when the uplink switching node receives the available bandwidth information, when forwarding to other downlink switching nodes, the available bandwidth information to be forwarded is updated according to the uplink available bandwidth of the outgoing port link, and the update method is: The minimum available bandwidth of the port link and the minimum value of the received available bandwidth; when the switching node A *.1 receives the available bandwidth information, it will receive the value as the available bandwidth of the path from its receiving end to the source switching node. .
  • each switching node A *.1 can obtain the available bandwidth information of the path to other peer switching nodes.
  • the switching node can also calculate the available bandwidth of the aggregation path to other switching nodes.
  • the available bandwidth of the aggregation path is equal to the available bandwidth of all parallel paths between the two switching nodes. Sum.
  • These switching nodes store these path information in their path information tables. For example, the path information tables of A 1.1 , A 2.1, and A 3.1 are shown in Table 1, Table 2, and Table 3, respectively. They are consistent with the results of Method 1 and Method 2.
  • the measurement of the aggregate available bandwidth of the congestion measurement path of the aggregation path between the peer switching nodes of distance 2 is described above with reference to FIG. 8 to FIG. 11; the same level exchange with the distance of 2k (k>1) will be described below with reference to FIG. Measurement of the aggregate available bandwidth of the congestion measurement path of the aggregation path between nodes.
  • the method can include two three implementations.
  • the first implementation manner may be that the switching node calculates the aggregate available bandwidth of the path between the upper port of the downlink switching node and the path to the other switching node that is in the same level as the downlink switching node, and sends the determined aggregate available bandwidth. Give the corresponding downstream switching node.
  • the second execution mode and the third execution mode are aggregated available bandwidth of the path between the switching node and its own to other switching nodes.
  • the basic steps of the first implementation are as follows:
  • Step a Each intermediate switching node (referred to as a "source intermediate switching node” for convenience of description) transmits to the other intermediate switching nodes that are reachable at the same level and carries each of its lower ports to its lower switching nodes. A message of the downlink available bandwidth information of the link.
  • Step b After receiving the message, the other intermediate switching node according to the available bandwidth information in the message, the aggregate available bandwidth to the source intermediate switching node, and the uplink available bandwidth information of the link from the lower port to the downlink switching node thereof, Calculate the aggregate available bandwidth of each of the downstream switching nodes of each of the downstream switching nodes to the source intermediate switching node.
  • the calculation method is to take the minimum value of the available bandwidth of each segment serial path. Then, the calculated aggregate available bandwidth is sent to the corresponding downlink switching node.
  • Step c After receiving the aggregate available bandwidth, the downlink switching node of the other intermediate switching node uses it as the aggregate available bandwidth of the uplink port of the source intermediate switching node.
  • the second implementation when other intermediate switching nodes receive the message carrying the available bandwidth information from the peer intermediate switching node, only the downlink switching to the source intermediate switching node may be calculated.
  • the intermediate switching node then sends the calculated aggregated available bandwidth to its downstream switching node, and the downlink switching node according to the received aggregated available bandwidth information and the available bandwidth of the uplink of the connected link of its receiving port, Calculate the aggregate available bandwidth of each of the downstream switching nodes on the upper port to the source intermediate switching node.
  • the calculation method is to take the minimum value of the available bandwidth of each segment serial path.
  • Step a Each intermediate switching node (referred to as a "source intermediate switching node” for convenience of description) sends a chain carrying its lower port to its downstream switching node to other intermediate switching nodes that are reachable at the same level.
  • the message of the downlink available bandwidth information of the road referred to as a "source intermediate switching node” for convenience of description
  • Step b After receiving the message of the source intermediate switching node, the other intermediate switching node calculates each downlink switching to the source intermediate switching node according to the available bandwidth information in the message and the aggregate available bandwidth information of the source intermediate switching node.
  • the calculation method is to take the minimum value of the available bandwidth of each segment serial path. The calculated aggregate available bandwidth is then sent to all of the downstream switching nodes.
  • Step c After receiving the aggregate available bandwidth, the downlink switching node calculates the uplink port to the other peer switching node according to the received aggregated available bandwidth and the available bandwidth of the uplink of the link connected to the receiving port. Aggregate available bandwidth. The calculation result is used as the aggregate available bandwidth of the upper port that receives the information to other peer switching nodes. The calculation method is to take the minimum value of the available bandwidth of each segment serial path.
  • the bandwidth information in the message may be sent to the source.
  • the aggregated bandwidth information of the switching node is sent to its downstream switching node. Calculating, by the downlink switching node, the downlink connection node of the upper port to the source intermediate switching node according to the received bandwidth information and the available bandwidth of the uplink of the link connected to the upper port of the bandwidth information. Aggregate available bandwidth information.
  • Step a Each intermediate switching node (referred to as a "source intermediate switching node” for convenience of description) sends a chain carrying its lower port to its downstream switching node to other intermediate switching nodes that are reachable at the same level.
  • the message of the downlink available bandwidth information of the road referred to as a "source intermediate switching node” for convenience of description
  • Step b After receiving the message of the source intermediate switching node, the other intermediate switching node obtains the bandwidth information in the message (that is, the downlink of the link of the source intermediate switching node to the downlink of the connected switching node is available)
  • the bandwidth information (that is, the downlink of the link of the source intermediate switching node to the downlink of the connected switching node is available)
  • the bandwidth information is sent to its downstream switching node.
  • Step c After the downlink switching node receives the bandwidth information sent by the intermediate switching node, according to the aggregate available bandwidth information of the intermediate switching node to the source intermediate switching node, the source intermediate switching node The downlink available bandwidth information of the link from the port to the connected switching node, and the available bandwidth of the uplink of the link connected to the upper port of the received bandwidth information, and calculating the upper port to each of the source intermediate switching nodes The aggregate available bandwidth of the downstream switching node.
  • the calculation method is to take the minimum value of the available bandwidth of each segment serial path. Then, the calculation result is used as the aggregate available bandwidth of the upper port to the lower switching node of the source intermediate switching node.
  • the downlink switching node can obtain the aggregate available bandwidth of each of the upper ports to other peer switching nodes.
  • the downlink switching node sums the aggregate available bandwidth from different upper ports to some other peer switching nodes, and can obtain the aggregate available bandwidth between the other switching nodes.
  • the downstream switching node is already an edge switching node, it may not be calculated to other peer switching sections. The aggregated bandwidth available for the point.
  • a 2.1 the intermediate switching node to switching node message carrying A 3.1 A 1.1 and that in connection link L ⁇ A 2.1 ,T 2.1> available downlink bandwidth information (value 4) and a lower link connected L ⁇ A 2.1 ,T 2.2>
  • the intermediate switching node A 1.1 receives the message, it calculates the aggregate available bandwidth of each of the downstream switching nodes of each of the downstream switching nodes to the lower switching nodes of A 2.1 .
  • the calculation method is described by taking the calculation of the available bandwidth of the port P3 of the switching node T 1.2 to the switching node T 2.2 as follows: taking the “downlink available bandwidth (value 3) of the link L ⁇ A 2.1 , T 2.2 > in the message” , the minimum value "polymerizable available bandwidth path information table in a 1.1 a 1.1 a 2.1 of the (value 4)" and "link L ⁇ A 1.1 ,T 1.2> available upstream bandwidth (value 2)" (the value 2 As the aggregate available bandwidth of the upper port P3 of the switching node T 1.2 to the switching node T 2.2 , that is, the minimum value of the available bandwidth of each segment is taken as the available bandwidth of the entire path.
  • a 1.1 can calculate the aggregate available bandwidth of each of the downstream switching nodes (T 2.1 and T 2.2 ) of the upper ports P3 to A 2.1 of its respective switching nodes (T 1.1 and T 1.2 ). Then A 1.1 sends the calculated aggregate available bandwidth information of each path to the source switching node of the path (the lower switching node of A 1.1 ). For example, A 1.1 sends the aggregate available bandwidth (value 2) of T 1.2 to T 2.1 and the aggregate available bandwidth (value 2) of T 1.2 to T 2.2 to T 1.2 .
  • the switching nodes T 1.1 , T 1.2 , T 3.1 and T 3.2 can obtain aggregated available bandwidth from their upper port P3 to switching nodes T 2.1 and T 2.2 , as shown in Table 3, Table 4, Table 5 and Table 12 6 is shown.
  • the switching node T *.* adds the aggregate available bandwidth of each of the upper ports to a switching node, and can further calculate the aggregate available bandwidth of the aggregation path between the switching nodes.
  • the execution of the second execution mode and the third execution mode are similar to those of the first execution mode. Only the switch nodes that calculate the available bandwidth of the aggregation are different, and are not described again.
  • the first switching node in the CLOS switching network may calculate the aggregate available bandwidth of the path between the first target port of the at least one lower level node connected to the next lower level node, so as to facilitate
  • the primary switching node dynamically schedules traffic between ports according to the aggregate available bandwidth of each port to implement load balancing. Improve the efficiency of measuring network paths.
  • FIG. 13 the prior art also has a variant of a multi-stage CLOS switching network that does not meet the definition of a standard CLOS switching network.
  • the root switching node is not only connected to one port of one switching node in each pod.
  • FIG. 14 is a standard multi-level CLOS switching network obtained by re-dividing the multi-stage CLOS switching network of FIG. By dividing its ports, we can always turn it into a logical standard multi-level CLOS network. Therefore, the embodiments of the present invention are equally applicable to a variant switching network of a CLOS switching network.
  • Figure 15 shows a schematic block diagram of an apparatus 1500 in accordance with an embodiment of the present invention.
  • Device 1500 can be a switch, or device 1500 can also be a physical module with switch functionality.
  • Apparatus 1500 can perform the steps performed by the first switching node in the methods of FIGS. 1 through 14.
  • the device 1500 includes: a communication module 1510 and a processing module 1520,
  • the processing module 1520 is configured to acquire, by using the communication module 1510, a first aggregate available bandwidth of a path of the first switching node to a second switching node, where the first switching node and the second switching node are a peer switching node in a CLOS switching network;
  • the processing module 1520 is further configured to acquire, by using the communication module 1510, a first available bandwidth of a path of the first target port of the third switching node to the first switching node, where the third switching node is a next-level switching node connected by a switching node;
  • the processing module 1520 is further configured to acquire, by using the communication module 1510, a second available bandwidth of a path of the second switching node to a fourth switching node, where the fourth switching node is connected to the second switching node.
  • Next level exchange node
  • the processing module 1520 is further configured to determine a second aggregate available bandwidth of the path of the first target port of the third switching node to the fourth switching node, where the second aggregate available bandwidth is available for the first aggregation The bandwidth, the first available bandwidth, and the smallest available bandwidth of the second available bandwidth.
  • the first switching node in the CLOS switching network may calculate, by using the foregoing manner, the aggregation of the path between the first target port of the at least one lower-level switching node connected to the next lower-level switching node. Bandwidth, so that the next-level switching node dynamically schedules traffic between ports according to the aggregate available bandwidth of each port to implement load balancing. This method reduces the complexity of measuring the network path and improves the efficiency of measuring the network path.
  • Figure 16 shows a schematic block diagram of an apparatus 1600 of an embodiment of the present invention.
  • Device 1600 can be a switch, or device 1600 can also be a physical module with switch functionality.
  • Apparatus 1600 can perform the steps performed by the third switching node in the methods of FIGS. 1 through 14.
  • the device 1600 includes: a communication module 1610 and a processing module 1620,
  • the processing module 1610 is configured to acquire, by using the communication module 1620, a first available bandwidth of a path of the first target port of the third switching node to the first switching node, where the third switching node is the first The next-level switching node connected by the switching node;
  • the processing module 1610 is further configured to acquire, by using the communication module 1620, a third aggregate available bandwidth of the path from the first switching node to the fourth switching node, where the fourth switching node is the same as the third switching node.
  • Level switching node
  • the processing module 1610 is further configured to determine a second aggregate available bandwidth of a path of the first target port of the third switching node to the fourth switching node, where the second aggregate available bandwidth is the first available bandwidth. And a smaller available bandwidth of the third aggregate available bandwidth.
  • the third switching node in the CLOS switching network may calculate the aggregate available bandwidth between the first target port and the fourth switching node in the foregoing manner. Therefore, the third switching node dynamically schedules traffic between the ports according to the aggregate available bandwidth corresponding to each port to implement load balancing. Improve the efficiency of measuring network paths.
  • Figure 17 shows a schematic block diagram of an apparatus 1700 in accordance with an embodiment of the present invention.
  • Device 1700 can be a switch, or device 1700 can also be a physical module with switch functionality.
  • Apparatus 1700 can perform the steps performed by the first switching node in the methods of FIGS. 1-14.
  • Apparatus 1700 includes:
  • a memory 1710 configured to store a program
  • a communication interface 1720 configured to communicate with other devices
  • the processor 1730 is configured to execute a program in the memory 1710. When the program is executed, the processor 1730 is configured to acquire, by using the communication interface 1710, a path of the path from the first switching node to the second switching node. Aggregating available bandwidth, the first switching node and the second switching node are peer switching nodes in the CLOS switching network; and acquiring, by the communication interface 1710, a first target port of the third switching node a first available bandwidth of a path of the first switching node, the third switching node is a next-level switching node connected to the first switching node; and acquiring, by the communication interface 1710, the second switching node to a second available bandwidth of the path of the fourth switching node, the fourth switching node is a next-level switching node connected to the second switching node; and determining a first target port of the third switching node to the a second aggregate available bandwidth of the path of the fourth switching node, the second aggregate available bandwidth being the first aggregate available bandwidth, the first available bandwidth, and the second available
  • the first switching node in the CLOS switching network may calculate and The aggregate available bandwidth of the path between the first target port of the at least one lower-level switching node and the next lower-level switching node, so that the next-level switching node can allocate the bandwidth according to the aggregated bandwidth of each port. Dynamically schedule traffic to achieve load balancing. This method reduces the complexity of measuring the network path and improves the efficiency of measuring the network path.
  • FIG. 18 shows a schematic block diagram of an apparatus 1800 in accordance with an embodiment of the present invention.
  • Device 1800 can be a switch, or device 1800 can also be a physical module with switch functionality.
  • Apparatus 1800 can perform the steps performed by the third switching node in the methods of FIGS. 1 through 14.
  • Apparatus 1800 includes:
  • a memory 1810 configured to store a program
  • a communication interface 1820 configured to communicate with other devices
  • the processor 1830 is configured to execute a program stored in the memory 1810. When the program is executed, the processor 1830 is configured to acquire, by using the communication interface 1820, the first target port of the third switching node to the first exchange. a first available bandwidth of a path of the node, the third switching node is a next-level switching node connected to the first switching node; and acquiring the first switching node to the fourth switching node by using the communication interface 1820 Third aggregate available bandwidth of the path, the fourth switching node and the third switching node are peer switching nodes; and determining a path of the first switching port of the third switching node to the fourth switching node The second aggregate available bandwidth, the second aggregate available bandwidth being the smaller of the first available bandwidth and the third aggregate available bandwidth.
  • the third switching node in the CLOS switching network may calculate the aggregate available bandwidth between the first target port and the fourth switching node in the foregoing manner. Therefore, the third switching node dynamically schedules traffic between the ports according to the aggregate available bandwidth corresponding to each port to implement load balancing. Improve the efficiency of measuring network paths.
  • system and “network” are used interchangeably herein.
  • the term “and/or” in this context is merely an association describing the associated object, indicating that there may be three relationships, for example, A and / or B, which may indicate that A exists separately, and both A and B exist, respectively. B these three situations.
  • the character "/" in this article generally indicates that the contextual object is an "or" relationship.
  • B corresponding to A means that B is associated with A, and B can be determined according to A.
  • determining B according to A does not mean that B is only determined based on A, and that B can also be determined based on A and/or other information.
  • the size of the sequence numbers of the above processes does not mean the order of execution, and the order of execution of each process should be determined by its function and internal logic, and should not be taken to the embodiments of the present invention.
  • the implementation process constitutes any limitation.
  • the processor in the embodiment of the present invention may be an integrated circuit chip with signal processing capability.
  • each step of the foregoing method embodiment may be completed by an integrated logic circuit of hardware in a processor or an instruction in a form of software.
  • the processor may be a general-purpose processor, a digital signal processor (DSP), an application specific integrated circuit (ASIC), a Field Programmable Gate Array (FPGA), or the like. Programming logic devices, discrete gates or transistor logic devices, discrete hardware components.
  • the methods, steps, and logical block diagrams disclosed in the embodiments of the present invention may be implemented or carried out.
  • the general purpose processor may be a microprocessor or the processor or any conventional processor or the like.
  • the steps of the method disclosed in the embodiments of the present invention may be directly implemented by the hardware decoding processor, or may be performed by a combination of hardware and software modules in the decoding processor.
  • the software module can be located in a conventional storage medium such as random access memory, flash memory, read only memory, programmable read only memory or electrically erasable programmable memory, registers, and the like.
  • the storage medium is located in the memory, the processor reads the information in the memory, and completes the above method in combination with the hardware thereof. A step of.
  • the memory in the embodiments of the present invention may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memory.
  • the non-volatile memory may be a read-only memory (ROM), a programmable read only memory (PROM), an erasable programmable read only memory (Erasable PROM, EPROM), or an electric Erase programmable read only memory (EEPROM) or flash memory.
  • the volatile memory can be a Random Access Memory (RAM) that acts as an external cache.
  • RAM Random Access Memory
  • many forms of RAM are available, such as static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (Synchronous DRAM).
  • SDRAM Double Data Rate SDRAM
  • DDR SDRAM Double Data Rate SDRAM
  • ESDRAM Enhanced Synchronous Dynamic Random Access Memory
  • SLDRAM Synchronous Connection Dynamic Random Access Memory
  • DR RAM direct memory bus random access memory
  • the disclosed system, apparatus, and method may be implemented in other manners.
  • the device embodiments described above are merely illustrative.
  • the division of the unit is only a logical function division.
  • there may be another division manner for example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored or not executed.
  • the mutual coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interface, device or unit, and may be electrical, mechanical or otherwise.
  • the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of the embodiment.
  • each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.
  • the functions may be stored in a computer readable storage medium if implemented in the form of a software functional unit and sold or used as a standalone product.
  • the technical solution of the present invention which is essential or contributes to the prior art, or a part of the technical solution, may be embodied in the form of a software product, which is stored in a storage medium, including
  • the instructions are used to cause a computer device (which may be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention.
  • the foregoing storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk, and the like, which can store program codes. .

Landscapes

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

Abstract

一种用于测量网络路径的方法、装置和系统，能够提高测量网络路径的效率。包括：CLOS交换网络中的第一交换节点获取第一交换节点到第二交换节点的路径的第一聚合可用带宽；第一交换节点获取第三交换节点的第一目标端口到第一交换节点的路径的第一可用带宽，第三交换节点是与第一交换节点连接的下一级交换节点；第一交换节点获取第二交换节点到第四交换节点的路径的第二可用带宽，第四交换节点是与第二交换节点连接的下一级交换节点；第一交换节点确定第三交换节点的第一目标端口到第四交换节点的路径的第二聚合可用带宽，第二聚合可用带宽为第一聚合可用带宽、第一可用带宽以及第二可用带宽中的最小可用带宽。

Description

用于测量网络路径的方法、装置和系统 技术领域
本发明涉及网络技术领域，尤其涉及用于测量网络路径的方法、装置和系统。
背景技术
随着云计算的发展，数据中心被大量地部署。CLOS交换网络是一种特殊的多级互联网络。由于其高性能和高可扩展，多级CLOS交换网络正逐步成为数据中心网络的主流架构。例如，图1和图2分别给出了两种不同级数的CLOS交换网络的示意图：叶脊(Spine-Leaf)交换网络和胖树(Fat-Tree)交换网络。其中，图1和图2中的圆代表计算节点，矩形表示交换节点，直线表示节点之间的链路。
如图1或图2所示，在多级CLOS交换网络中，一对边缘交换节点间存在多条路径，为了实现网络的高性能，需要在不同的路径间执行有效的负载均衡。即基于路径的拥塞状况，动态地调度流量，以使得网络的负载分布均匀，减少网络局部拥塞，实现网络的高性能。但是要快速、准确和低开销地获得路径拥塞状况是比较困难的。逐级反压法是现有常用的一种拥塞测量方法。在逐级反压法中，交换节点之间沿各路径逐级反向发送拥塞信息，从而通过拥塞信息确定一对边缘交换节点之间的各路径的拥塞状况。其中，拥塞信息可以有多种表示方法，比如路径的可用带宽。但是基于逐级反压法的网络路径测量方法不够精确。
发明内容
本发明提供了一种用于测量网络路径的方法，装置和系统，能够提高测量网络路径的效率。
第一方面，提供了一种用于测量网络路径的方法，包括：CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，所述第一交换节点和所述第二交换节点为所述CLOS交换网络中的同级交换节点；所述第一交换节点获取第三交换节点的第一目标端口到所述第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；所述第一交换节点获取所述第二交换节点到第四交换节点的路径的第二可用带宽，所述第四交换节点是与所述第二交换节点连接的下一级交换节点；所述第一交换节点确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一聚合可用带宽、所述第一可用带宽以及所述第二可用带宽中的最小可用带宽。
CLOS交换网络中的第一交换节点可以通过上述方式计算与其连接的至少一个下一级交换节点的第一目标端口到另一下一级交换节点之间的路径的聚合可用带宽，以便于下一级交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。该方法减少了测量网络路径的复杂度，提高了测量网络路径的效率。
在一种可能的实现方式中，所述CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，包括：所述第一交换节点分别从多个第 五交换节点接收第一信息，所述第一信息指示所述多个第五交换节点中的每个第五交换节点对应的第一路径的聚合可用带宽，其中，所述多个第五交换节点是与所述第一交换节点连接的上一级交换节点，所述第一路径是经过所述每个第五交换节点的所述第一交换节点到所述第二交换节点的路径；所述第一交换节点根据所述第一信息，将所述多个第五交换节点对应的第一路径的聚合可用带宽之和确定为所述第一聚合可用带宽。
在一种可能的实现方式中，所述第一交换节点和所述第二交换节点是与所述CLOS交换网络中的多个第六交换节点连接的下一级交换节点，所述CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，包括：所述第一交换节点获取所述第一交换节点到所述多个第六交换节点中的每个第六交换节点的路径的上行可用带宽；所述第一交换节点从所述每个第六交换节点接收第二信息，所述第二信息用于指示所述每个第六交换节点到所述第二交换节点的下行可用带宽；所述第一交换节点确定所述上行可用带宽和所述下行可用带宽中较小的可用带宽为所述每个第六交换节点对应的可用带宽；将所述多个第六交换节点对应的可用带宽之和确定为所述第一聚合可用带宽。
在一种可能的实现方式中，所述第一交换节点获取所述第二交换节点到第四交换节点的路径的第二可用带宽，包括：所述第一交换节点接收所述第二交换节点发送的第三信息，所述第三信息用于指示所述第二交换节点到第四交换节点的路径的可用带宽是所述第二可用带宽。
在一种可能的实现方式中，还包括：所述第一交换节点向所述第三交换节点发送第四信息，所述第四信息用于指示所述第三交换节点的第一目标端口到所述第四交换节点之间的可用带宽是所述第二聚合可用带宽。
第二方面，提供了一种用于测量网络路径的方法，包括：CLOS交换网络中的第三交换节点获取所述第三交换节点的第一目标端口到第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；所述第三交换节点获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，所述第四交换节点与所述第三交换节点为同级交换节点；所述第三交换节点确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一可用带宽和所述第三聚合可用带宽中较小的可用带宽。
CLOS交换网络中的第三交换节点可以通过上述方式计算其第一目标端口到第四交换节点之间的聚合可用带宽。以便于第三交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。提高了测量网络路径的效率。
在一种可能的实现方式中，所述第三聚合可用带宽是第一聚合可用带宽和第二可用带宽中的较小的可用带宽，所述第一聚合可用带宽是所述第一交换节点到第二交换节点的路径的聚合可用带宽，所述第二可用带宽是所述第二交换节点到所述第四交换节点的路径的第二可用带宽，所述第二交换节点是与所述第四交换节点连接的上一级交换节点。
在一种可能的实现方式中，所述第三交换节点获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，包括：所述第三交换节点从所述第一交换节点接收第一指示信息，所述第一指示信息用于指示所述第一交换节点到所述第四交换节点的路径的聚合可用带宽为所述第三聚合可用带宽。
在一种可能的实现方式中，所述第三交换节点获取所述第一交换节点到第四交换节点 的路径的第三聚合可用带宽，包括：所述第三交换节点接收所述第一交换节点发送的第二指示信息，所述第二指示信息用于指示所述第一交换节点到第二交换节点的路径的第一聚合可用带宽；所述第三交换节点接收所述第二交换节点发送的第三指示信息，所述第三指示信息用于指示所述第二交换节点到所述第四交换节点的路径的第二可用带宽；所述第三交换节点确定所述第三聚合可用带宽，所述第三聚合可用带宽是第一聚合可用带宽和第二可用带宽中的较小的可用带宽。
在一种可能的实现方式中，还包括：所述第三交换节点确定所述第三交换节点到所述第四交换节点的第四聚合可用带宽，所述第四聚合可用带宽为所述第三交换节点的各端口分别到所述第四交换节点的路径的聚合可用带宽之和。
第三方面，提供了一种装置，所述装置包括用于执行第一方面的方法的模块。基于同一发明构思，由于该装置解决问题的原理与第一方面的方法设计中的方案对应，因此该装置的实施可以参见方法的实施，重复之处不再赘述。
第四方面，提供了一种装置，所述装置包括用于执行第二方面的方法的模块。基于同一发明构思，由于该装置解决问题的原理与第二方面的方法设计中的方案对应，因此该装置的实施可以参见方法的实施，重复之处不再赘述。
第五方面，提供了一种交换网络系统，所述交换网络系统包括上述第三方面的装置以及上述第四方面的装置。
第六方面，提供了一种装置，包括存储器，用于存储程序；收发器，用于和其他设备进行通信；处理器，用于执行存储器中的程序，当所述程序被执行时，当所述程序被执行时，所述处理器用于执行第一方面的方法。
第七方面，提供了一种装置，包括存储器，用于存储程序；收发器，用于和其他设备进行通信；处理器，用于执行存储器中的程序，当所述程序被执行时，当所述程序被执行时，所述处理器用于执行第二方面的方法。
第八方面，提供了一种交换网络系统，所述交换网络系统包括上述第六方面的装置以及上述第七方面的装置。
第九方面，提供了一种系统芯片，所述系统芯片包括用于执行第一方面的方法的模块。
第十方面，提供了一种系统芯片，所述系统芯片包括用于执行第二方面的方法的模块。
第十一方面，提供了一种交换网络系统，所述交换网络系统包括上述第九方面的系统芯片以及上述第十方面的系统芯片。
附图说明
为了更清楚地说明本发明实施例的技术方案，下面将对本发明实施例中所需要使用的附图作简单地介绍，显而易见地，下面所描述的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲，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他的附图。
图1是本发明实施例的Spine-Leaf交换网络的结构示意图；
图2是本发明实施例的Fat-Tree交换网络的结构示意图；
图3是本发明实施例的基于反压的测量网络路径的方法的示意图；
图4是本发明另一实施例的基于反压的测量网络路径的方法的示意图；
图5是本发明另一实施例的多级CLOS交换网络的结构示意图；
图6是本发明另一实施例的用于测量网络路径的方法的示意性框图；
图7是本发明再一实施例的用于测量网络路径的方法的示意性框图；
图8是本发明再一实施例的CLOS交换子网络的拓扑示意图；
图9是本发明又一实施例的用于测量网络路径的方法的示意图；
图10是本发明又一实施例的用于测量网络路径的方法的示意图；
图11是本发明又一实施例的用于测量网络路径的方法的示意图；
图12是本发明又一实施例的用于测量网络路径的方法的示意图；
图13是本发明又一实施例的多级CLOS网络的结构示意图；
图14是本发明又一实施例的多级CLOS网络的结构示意图；
图15是本发明实施例的装置的示意图；
图16是本发明另一实施例的装置的示意图；
图17是本发明另一实施例的装置的示意图；
图18是本发明另一实施例的装置的示意图。
具体实施方式
为使本发明实施例的目的、技术方案和优点更加清楚，下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清楚、完整地描述，显然，所描述的实施例是本发明一部分实施例，而不是全部的实施例。基于本发明中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本发明保护的范围。
需要说明的是，本发明实施例中的带宽可以采用绝对带宽值表示，也可以采用量化值表示。例如，带宽可以用位/秒(bit/s，即bps)表示，也可以用带宽量化值表示，本发明实施例对带宽的表示方式不作限制。
另外，在本发明实施例中，测量网络路径通常可以指测量交换节点之间的最短网络路径。
为了便于理解本发明实施例，首先介绍CLOS交换网络的概念和架构。
CLOS交换网络是一种特殊的多级互连网络(Multistage Interconnection Networks)。例如，CLOS交换网络可以包括图1所示的Spine-Leaf交换网络和/或图2所示的Fat-Tree交换网络。作为一个示例，本发明中所述多级CLOS交换网络可以通过下述方法构造得到：取两组交换节点，让第一组中的每个交换节点连接第二组中的每个交换节点，可以构造一个二级CLOS交换网络。将该二级CLOS交换网络看做一个虚拟交换节点，取多个虚拟交换节点构成一组，并让每个虚拟交换节点连接到第三组交换节点中的每个交换节点，可以构成一个三级CLOS交换网络。例如，如图2所示，每个虚拟交换节点可以称为区块(pod)。递归地，将一个i级CLOS交换网络看作一个虚拟交换节点，取多个虚拟交换节点构成一组，并让每个虚拟交换节点连接到第i+1组交换节点中的每个交换节点，可以构成一个i+1级CLOS交换网络。
在一个i+1级CLOS交换网络中，将第一组交换节点中的交换节点称为边缘交换节点，将第i+1组换节点中的交换节点称为核心交换节点或根交换节点，将第二组到第i组交换节点中的交换节点称为中间交换节点。特殊地，对于三级CLOS交换网络，在本发明实施例中可以将边缘交换节点称为TOR交换节点(用Ti.j表示)，将核心交换节点称为Core交换节点(用Ci.j表示)，将中间交换节点称为Agg交换节点(用Ai.j表示)。
在一个i+1级CLOS交换网络中，仅边缘交换节点连接计算节点(例如，服务器)。
为了便于理解本发明实施例，下面介绍现有技术中基于逐级反压法测量网络路径的方法。例如，图3显示了如何基于逐级反压实现边缘交换节点T1.1到T2.1的多条路径的拥塞的量。在图3中，用可用带宽的量化值来度量拥塞状况，值越大表示可用带宽越大，拥塞程度越低。每条链路用“(链路上行可用带宽，链路下行可用带宽)”来表示链路两个方向上的可用带宽。结合图3示例，对该测量过程进行说明：首先，交换节点T2.1通过其上端口P3(需要说明的是，本发明实施例中的P*用于指示对应交换节点的*端口，例如，P3表示T2.1的3端口)和P4分别向上一跳交换节点A2.1和A2.2发送指示信息，以指示到达上一跳交换节点到达T2.1的路径的可用带宽；其次，当某交换节点，比如A2.1，接收到指示信息后，对该指示信息进行存储，并进一步向上一跳交换节点C1.1和C1.2反馈可用带宽信息。此时反馈的可用带宽信息需要根据A2.1与上一跳交换节点间的链路的可用带宽信息进行更新，其更新方法是：取接收收到的可用带宽值和链路的可用带宽的最小值。即将路径上瓶颈链路的可用带宽作为路径的可用带宽。该过程可以一直进行，直到可用带宽信息反馈到边缘交换节点T1.1。在反压过程中，指示信息中可记录路径信息，指示信息被最终反馈到边缘节点时，边缘节点可区别不同路径。如图3所示，交换节点T1.1将收到反压自交换节点T2.1的四条不同路径的可用带宽信息。图4显示了这四条不同的路径及其可用带宽，可用带宽用圆圈内数字表示。基于该方法，可以测量任意两个交换节点间所有最短路径的可用带宽。
上述基于逐级反压的网络路径测量方法存在以下缺点：首先，由于两边缘交换节点间的路径数目可能非常多，传递拥塞信息会占用大量的链路带宽。比如：对于由m个端口的同构交换节点搭建的Fat-Tree网络，一对边缘交换节点间的不同路径数可高达m2/4条。其次，对于一对边缘交换节点间的多条路径，无法区分它们是否共享可用带宽。例如，如图4所示，从交换节点T1.1的上端口P4出发的路径3和路径4的可用带宽分别为3和1，其可用带宽之和为4。但是路径3和路径4的共享链路<T2.1,A2.2>的上行链路的可用带宽为3，所以路径3和路径4共享可用带宽。即这两条路径的聚合可用带宽小于该两条路径的可用带宽之和。而从交换节点T1.1的上端口P3出发的路径1和路径2的可用带宽分别为2和2，其可用带宽之和为4，它们共享物理链路<T1.1,A1.1>的可用带宽为4，所以路径1和路径2不共享可用带宽。即这两条路径的聚合可用带宽等于它们的可用带宽之和。
为了减少测量网络路径占用的链路带宽以及提高测量网络路径的效率，本发明实施例提出了一种用于测量网络路径的方法、装置和系统。该方案基于以下核心思想获得：
首先，对于CLOS交换网络的路径测量，尽管一对边缘节点之间存在多条路径(例如，图4中的T1.1到T2.1存在4条路径)，但对于某一边缘交换节点来说，没必要知道如何在这多条路径间分配流量，而只需知道它的各个出端口(例如，图4中的T1.1的端口P3和端口P4)之间各应分配多少流量。依次类推，每个交换节点都只需要关心如何在它的出端口上分配流量，而无需关心其到目的交换节点的各条路径上分别要分配多少流量。这使得单个交换节点需要维护的信息量从“其到目的交换节点的路径数”降为“其出端口数”。
其次，尽管交换节点(例如，图4中的T1.1)只需知道如何在它的各出端口上分配流量，但应分配的流量多少不应取决于其出端口所连链路有多少可用带宽，而应取决于从它的每个出端口到目的交换节点(例如，图4中T1.1到T2.1)各有多少聚合可用带宽。此处将源交换节点的某一出端口到目的交换节点的一组路径，看作一条聚合路径，该聚合路径 的可用带宽可以称作聚合可用带宽。并且，交换节点的某一出端口到另一交换节点的聚合可用带宽往往不等于从该出端口到另一交换节点的各路径的可用带宽之和，因为这多条路径间可能共享可用带宽。比如：图4中的T1.1的端口P3到T2.1的可用带宽不一定等于路径“T1.1->A1.1->C1.1->A2.1->T2.1”和路径“T1.1->A1.1->C1.2->A2.1->T2.1”的可用带宽之和，因为这两条路径可能共享链路L<T1.1,A1.1>和链路L<A2.1,T2.1>的可用带宽。一种正确的计算聚合路径的聚合可用带宽的方法是：将聚合路径切分为串并联路径段的组合，并联的各路径段的可用带宽求和，串联的各路径段求这些路径段的可用带宽的最小值。
所以，在本发明实施例中，将聚合路径切分为串并联路径段的组合，将聚合路径的聚合可用带宽的计算转化为对并联的各路径段的聚合可用带宽的计算和对串联的各路径段的聚合可用带宽计算的组合。其中，对并联的各路径段的聚合可用带宽的计算方法是求各路径段的(聚合)可用带宽之和，对串联的各路径段的聚合可用带宽的计算方法是求这些路径段的(聚合)可用带宽的最小值。
可选地，在本发明实施例中，对于并联的各路径段的聚合可用带宽的计算可以有多种定义，至少包括两种。其中，第一种定义可以是如上所述的对并联的各路径段的可用带宽求和。第二种定义可以是对并联的各路径段的可用带宽求平均值(即将并联的各路径段的可用带宽之和除以并联的路径数)。为了描述方便，本发明实施例中采用第一种定义。显然，在第二种定义的情况下，本发明实施例的测量网络路径的方法也可以实施，此处不再赘述。本领域技术人员根据聚合可用带宽的以上计算方式，显然可以进行各种等价的修改或变化，这样的修改或变化也落入本发明实施例的范围内。
本领域技术人员能够理解，并非任意网络中，源交换节点和目的交换节点间的聚合路径都能切分为串并联路径的组合，但是在CLOS交换网络中源交换节点和目的交换节点间的聚合路径具有这样的特性。图5示出了本发明实施例的CLOS交换网络的一对边缘节点间路径的示意图，其展示了多级CLOS交换网络的交换节点间路径从内到外分层聚合的特点。如图5所示，由于多级CLOS交换网络拓扑的规则性，CLOS交换网络中位于同一级的两个交换节点之间的路径存在从内到外分层聚合的特点。这能够简化路径的聚合可用带宽的测量。所以，多级CLOS交换网络的源边缘交换节点到目的边缘交换节点间的聚合路径能够切分成这样的串并联路径端的组合，并且具有嵌套特性，可以采用一种递归的方法求解聚合可用路径带宽。其主要思想是，可以先测量内层路径的拥塞状态(例如，通过可用带宽表征拥塞状态)，并进行汇聚，然后向外逐层扩展和汇聚，获得CLOS交换网络中各同级交换网络节点之间路径的聚合可用带宽信息。本发明实施例可以基于这种递归的方法获取CLOS交换网络的聚合可用带宽。
图6示出了本发明实施例的用于测量网络路径的方法的示意性框图。该方法600可以由CLOS交换网络中的交换节点执行。如图6所示，该方法600包括：
S601，CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，所述第一交换节点和所述第二交换节点为所述CLOS交换网络中的同级交换节点。
其中，该第一交换节点可以为CLOS交换网络中的中间交换节点。例如，第一交换节点可以是图4中的A1.1。上述第一聚合可用带宽可以是第一交换节点到第二交换节点之间的一组路径的聚合可用带宽。或者说，第一聚合可用带宽是第一交换节点到第二交换节点之间的所有路径的聚合可用带宽。或者说，第一聚合可用带宽是第一交换节点到第二交换 节点之间的聚合路径的实际可用带宽。
S602，所述第一交换节点获取第三交换节点的第一目标端口到所述第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点的连接的下一级交换节点。
可选地，所述第三交换节点是与所述第一交换节点相连的下一级交换节点。例如，第一交换节点可以是图4中的A1.1，第三交换节点可以是图4中的T1.1，所述第一目标端口可以是图4中T1.1的端口P3。其中，所述第三交换节点的第一目标端口到所述第一交换节点的路径可以是所述第三交换节点到所述第一交换节点的上行链路。
S603，所述第一交换节点获取所述第二交换节点到第四交换节点的路径的第二可用带宽，所述第四交换节点是与所述第二交换节点连接的下一级交换节点。
可选地，所述第四交换节点是与所述第二交换节点相连的下一级交换节点。例如，所述第二交换节点可以为图4中的A2.1,所述第四交换节点可以为图4中的T2.1。第二交换节点到第四交换节点的路径可以是第二交换节点到第四交换节点之间的下行链路。
S604，所述第一交换节点确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一聚合可用带宽、所述第一可用带宽以及所述第二可用带宽中的最小可用带宽。
可选地，该第二聚合可用带宽可以是第三交换节点的第一目标端口到第四交换节点的路径的聚合可用带宽。利用本发明实施例提供的用于测量网络路径的方法，CLOS交换网络中的交换节点可以获取其多个出端口中的每个出端口到另一交换节点之间的聚合可用带宽，以便于交换节点在各端口之间动态调度流量，执行负载均衡。
在本发明实施例中，CLOS交换网络中的第一交换节点可以通过上述方式计算与其连接的至少一个下一级交换节点的第一目标端口到另一下一级交换节点之间的路径的聚合可用带宽，以便于下一级交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。该方法减少了测量网络路径的复杂度，提高了测量网络路径的效率。
可选地，本发明实施例中，可以采用周期性或事件触发的方式执行网络路径的测量。
可选地，在本发明实施例中，第一交换节点可以接收其上一级交换节点发送的信息，以指示第一交换节点与上一级节点连接的出端口到第二交换节点之间的聚合可用带宽。第一交换节点可以根据多个出端口对应的聚合可用带宽之和，确定第一交换节点与第二交换节点的路径的第一聚合可用带宽。
例如，所述CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，包括：所述第一交换节点分别从多个第五交换节点接收第一信息，所述第一信息指示所述多个第五交换节点中的每个第五交换节点对应的第一路径的聚合可用带宽，其中，所述多个第五交换节点是与所述第一交换节点连接的上一级交换节点，所述第一路径是经过所述每个第五交换节点的所述第一交换节点到所述第二交换节点的路径；所述第一交换节点根据所述第一信息，将所述多个第五交换节点对应的第一路径的聚合可用带宽之和确定为所述第一聚合可用带宽。
其中，上述第一路径可以是第一交换节点到第二交换节点的一条路径或一组路径。第一路径的个数取决于经过每个第五交换节点的、第一交换节点到第二交换节点的路径的个数。换句话说，第一路径是第一交换节点的第一目标端口到第二交换节点之间的路径。第 一交换节点的第一目标端口可以是第一交换节点到第五交换节点的路径所经过的出端口。另外，第五交换节点可以是CLOS交换网络中的根交换节点，也可以是CLOS交换网络中的中间交换节点。当第五交换节点为中间交换节点的情况下，每个第五交换节点对应的第一路径可以包括多条路径。当第五交换节点为根交换节点的情况下，每个第五交换节点对应的第一路径可以是一条路径。
可选地，当第一交换节点和第二交换节点连接相同的上一级交换节点的情况下，假设第一交换节点和第二交换节点同时连接的上一级交换节点为第六交换节点，第一交换节点到第二交换节点之间的第一聚合带宽可以采用以下方法确定：
所述第一交换节点和所述第二交换节点是与所述CLOS交换网络中的多个第六交换节点连接的下一级交换节点，所述CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，包括：所述第一交换节点获取所述第一交换节点到所述多个第六交换节点中的每个第六交换节点的路径的上行可用带宽；所述第一交换节点从所述每个第六交换节点接收第二信息，所述第二信息用于指示所述每个第六交换节点到所述第二交换节点的下行可用带宽；所述第一交换节点确定所述上行可用带宽和所述下行可用带宽中较小的可用带宽为所述每个第六交换节点对应的可用带宽；将所述多个第六交换节点对应的可用带宽之和确定为所述第一聚合可用带宽。
可选地，在本发明实施例中，所述第一交换节点获取所述第二交换节点到第四交换节点的路径的第二可用带宽，包括：所述第一交换节点接收所述第二交换节点发送的第三信息，所述第三信息用于指示所述第二交换节点到第四交换节点的路径的可用带宽是所述第二可用带宽。
可选地，在本发明实施例中，第一交换节点在确定与其相连的第三交换节点的第二聚合可用带宽之后，可以向第三交换节点发送消息，指示第三交换节点的第一目标端口到其他交换节点之间的聚合可用带宽。以便于第三交换节点储存信息并执行负载均衡。例如，方法600还包括：所述第一交换节点向所述第三交换节点发送第四信息，所述第四信息用于指示所述第三交换节点的第一目标端口到所述第四交换节点之间的可用带宽是所述第二聚合可用带宽。
方法600结合图6，描述了由上一级交换节点计算下一级交换节点的各端口之间的路径的聚合可用带宽的方法。图7示出了本发明另一实施例的用于测量网络路径的方法的示意性框图。下文将结合图7，描述本发明另一实施例的用于测量网络路径的方法700。该方法700可以由CLOS交换网络中的交换节点执行。方法700描述了由本级交换节点计算本级交换节点的各端口之间的路径的聚合可用带宽的方法。图7的方法中与图6的方法相同或相似的内容，请参考图6的相关部分，此处不再赘述。方法700包括：
S701，CLOS交换网络中的第三交换节点获取所述第三交换节点的第一目标端口到第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点。
其中，第三交换节点可以为CLOS交换网络中的中间交换节点或边缘交换节点。例如，第三交换节点可以为图4中的T1.1，第一交换节点可以为图4中的A1.1。
S702，所述第三交换节点获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，所述第四交换节点与所述第三交换节点为同级交换节点。
S703，所述第三交换节点确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一可用带宽和所述第三聚合可用带宽中较小的可用带宽。
在本发明实施例中，CLOS交换网络中的第三交换节点可以通过上述方式计算其第一目标端口到第四交换节点之间的聚合可用带宽。以便于第三交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。提高了测量网络路径的效率。
可选地，在方法700中，所述第三聚合可用带宽是第一聚合可用带宽和第二可用带宽中的较小的可用带宽，所述第一聚合可用带宽是所述第一交换节点到第二交换节点的路径的聚合可用带宽，所述第二可用带宽是所述第二交换节点到所述第四交换节点的路径的第二可用带宽，所述第二交换节点是与所述第四交换节点连接的上一级交换节点。
可选地，在方法700中，所述第三交换节点获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，包括：所述第三交换节点从所述第一交换节点接收第一指示信息，所述第一指示信息用于指示所述第一交换节点到所述第四交换节点的路径的聚合可用带宽为所述第三聚合可用带宽。
可选地，在方法700中，所述第三交换节点获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，包括：所述第三交换节点接收所述第一交换节点发送的第二指示信息，所述第二指示信息用于指示所述第一交换节点到第二交换节点的路径的第一聚合可用带宽；所述第三交换节点接收所述第二交换节点发送的第三指示信息，所述第三指示信息用于指示所述第二交换节点到所述第四交换节点的路径的第二可用带宽；所述第三交换节点确定所述第三聚合可用带宽，所述第三聚合可用带宽是第一聚合可用带宽和第二可用带宽中的较小的可用带宽。
可选地，在本发明实施例中，当第三交换节点获的第三交换节点的各出端口分别对应的到第四交换节点的路径的聚合可用带宽之后，第三交换节点还可以确定第三交换节点到第四交换节点的所有路径的聚合可用带宽。例如，在方法700中，还包括：所述第三交换节点确定所述第三交换节点到所述第四交换节点的第四聚合可用带宽，所述第四聚合可用带宽为所述第三交换节点的各端口分别到所述第四交换节点的路径的聚合可用带宽之和。
可选地，在方法700中，若第三交换节点并非边缘交换节点，则还存在与第三交换节点连接的下一级交换节点。第三交换节点可以向下一级交换节点指示第三交换节点到第四交换节点的聚合可用带宽，以便于下一级交换节点继续计算下一级交换节点之间的聚合可用带宽。例如，方法700还可以包括：所述第三交换节点向第七交换节点发送第四指示信息，所述第四指示信息用于指示所述第三交换节点和所述第四交换节点之间的可用带宽为所述第四聚合可用带宽，所述第七交换节点为所述第三交换节点的下一级交换节点。
上文结合图1至图7，介绍了本发明实施例的用于测量网络路径的方法，下文将结合图8至图12中的具体例子，更加详细地描述本发明实施例。应注意，图8至图12的例子仅仅是为了帮助本领域技术人员理解本发明实施例，而非要将本发明实施例限于所例示的具体数值或具体场景。本领域技术人员根据所给出的图8至图12的例子，显然可以进行各种等价的修改和变化，这样的修改和变化也落入本发明实施例的范围。
据多级CLOS网络的拓扑特征，本发明实施例的用于测量网络路径的方法可以分为两种类型：一种为，距离为2的同级交换节点间路径的聚合可用带宽的测量；另一种为，距离为2k(k>1)的同级交换节点间路径的聚合可用带宽的测量。此处两交换 节点间的距离是指从一个交换节点到另一个交换节点所经过的链路的数目。在实施中，上述两种类型的方法可并行执行。
下面对本发明的技术方案进行详细阐述，依次介绍如何执行距离为2的同级交换节点间路径的聚合可用带宽的测量，以及如何执行距离为2k(k>1)的同级交换节点间路径的聚合可用带宽的测量。
在多级CLOS网络中，由上一级交换节点互连的多个下一级交换节点间的距离为2。此处以图8为例，对如何执行距离为2的同级交换节点间路径的聚合可用带宽的测量进行说明。图8是图2所示Fat-Tree交换网络拓扑的一个子图，该图中各交换节点A*.1间的距离为2。图9至图11分别示出了测量图8中的同级交换节点间路径的聚合可用带宽的方法。如图9至图11所示，本发明实施例分别提出三种执行距离为2的同级交换节点间路径的聚合可用带宽的测量的方法。
在图9所示的方法中，可以由上一级交换节点集中式计算其下连的多个交换节点间的路径的可用带宽，然后将路径信息下发到对应下连交换节点，具体方法如图9所示。
如图9所示，每条链路用上下行两个方向上的可用带宽用“(链路上行可用带宽，链路下行可用带宽)”表示。交换节点C1.*通过测量其各端口发送/接收方向带宽占用情况可获得各链路的可用带宽信息。根据各链路的可用带宽信息，通过对连接两下连交换节点的链路的可用带宽取最小值可以计算得到两下连交换节点间路径的可用带宽。比如，交换节点C1.1通过链路L<C1.1,A1.1>连接交换节点A1.1的端口P3，通过链路L<C1.1,A2.1>连接交换节点A2.1的端口P3，通过取链路L<C1.1,A1.1>的上行可用带宽(值3)和链路L<C1.1,A2.1>的下行可用带宽(值2)的最小值，可以计算得到从A1.1的端口P3到A2.1的端口P3的路径的可用带宽为2。按此方法，交换节点C1.1和C1.2可分别计算得到其下连各交换节点A1.*间的路径信息表，如图9中表1和表2所示。
当交换节点C1.*通计算得到其下连各交换节点A*.1间的路径信息表后，将对应的路径信息(例如，该路径信息相当于方法600中的第一信息)发送给路径的源交换节点。因此各交换节点A*.1可以获得其到其他同级交换节点的路径可用带宽信息。同时，由于两个交换节点间可能有多条并行路径，交换节点还可以计算得到其到其他交换节点的聚合路径的可用带宽，聚合路径的可用带宽等于两个交换节点间所有并行路径的可用带宽之和。这些交换节点可以将这些路径信息存储在它们的路径信息表中。比如，A1.1、A2.1和A3.1的路径信息表分别如图9中表3、表4和表5所示。由图9中表4可知，从交换节点A2.1的端口P3和P4分别有一条路径到交换节点A3.1，它们的可用带宽分别为2和3，因此从交换节点A2.1到交换节点A3.1的聚合可用带宽为5。
在图10所示的方法中，交换节点(为了便于区分，可以称作“源交换节点”)可以向其上连交换节点所连的其他同级交换节点发送其上端口所连链路的下行可用带宽信息；当同级其他交换节点收到该可用带宽信息后，同级其他交换节点可以计算其到该下行可用带宽信息的源交换节点的路径信息。计算方法为：取接收可用带宽信息的端口所连链路的上行可用带宽与接收到的可用带宽值的最小值。比如，如图10所示，A3.1通过上端口P4向节点A1.1和A2.1发送链路L<C1.2,A2.1>的下行可用带宽(值6)；当A1.1收到该可用带宽信息后，将其与链路L<C1.2,A1.1>的上行可用带宽(值3)进行比较并取最小值(值3)，得到从A1.1的端口P4到A2.1的路径的可用带宽为3。按此 方法，各交换节点A*.1可以获得其到其他同级交换节点的路径可用带宽信息。同时，由于两个交换节点间可能有多条并行路径，交换节点还可以计算得到其到其他交换节点的聚合路径的聚合可用带宽，聚合路径的聚合可用带宽等于两个交换节点间所有并行路径的可用带宽之和。这些交换节点将这些路径信息存储在它们的路径信息表中。比如，A1.1、A2.1和A3.1的路径信息表分别如图10中表1、表2和表3所示。它们与图9中表3、表4和表5是一致的。
在图11所示的方法中，交换节点(为了便于区分，可以称作“源交换节点”)向由上连交换节点互连的其他同级交换节点发送其上端口所连链路的下行可用带宽信息；当上连交换节点收到该可用带宽信息后，在向其他下连交换节点转发时，根据出端口链路的上行可用带宽对需要转发的可用带宽信息进行更新，更新方法为：取出端口链路的上行可用带宽与接收到的可用带宽值的最小值；当交换节点A*.1收到可用带宽信息时，将收到的值作为其接收端到源交换节点的路径的可用带宽。比如，如图11所示，A3.1通过上端口P4向节点A1.1和A3.1发送链路L<C1.2,A2.1>的下行可用带宽(值6)；当C1.2向A1.1转发该可用带宽信息时，会将其与链路L<C1.2,A1.1>的上行可用带宽(值3)进行比较并取最小值(值3)，并用该值更新可用带宽信息，然后将更新后的可用带宽信息转发给交换节点A1.1；从A1.1的端口P4到收到可用带宽信息时，将该该可用带宽值(值3)作为其端口P4到交换节点A3.1的路径的可用带宽。按此方法，各交换节点A*.1可以获得其到其他同级交换节点的路径可用带宽信息。同时，由于两个交换节点间可能有多条并行路径，交换节点还可以计算得到其到其他交换节点的聚合路径的可用带宽，聚合路径的可用带宽等于两个交换节点间所有并行路径的可用带宽之和。这些交换节点将这些路径信息存储在它们的路径信息表中。比如，A1.1、A2.1和A3.1的路径信息表分别如图11中表1、表2和表3所示。它们与方法1和方法2的结果是一致的。
上文结合图8至图11介绍了距离为2的同级交换节点间聚合路径的拥塞测量路径的聚合可用带宽的测量；下文将结合图12介绍距离为2k(k>1)的同级交换节点间聚合路径的拥塞测量路径的聚合可用带宽的测量。该方法可以包括两种三种执行方式。第一种执行方式可以是交换节点计算其下连交换节点的上端口到与所述下连交换节点同级的到其他交换节点之间的路径的聚合可用带宽，并将确定的聚合可用带宽发送给相应的下连交换节点。第二种执行方式和第三种执行方式是交换节点计算自身到其他交换节点的之间的路径的聚合可用带宽。第一种执行方式的基本步骤如下：
步骤a.各中间交换节点(为了便于描述，将该中间交换节点称为“源中间交换节点”)向其他同级可达的中间交换节点发送携带其各下端口到其个下连交换节点的链路的下行可用带宽信息的消息。
步骤b.其他中间交换节点收到消息后，根据消息中的可用带宽信息、其到上述源中间交换节点的聚合可用带宽、其下端口到其下连交换节点的链路的上行可用带宽信息，计算其各下连交换节点到上述源中间交换节点的各下连交换节点的聚合可用带宽。计算方法是取各段串联路径的可用带宽的最小值。然后，将计算得到的聚合可用带宽发送给对应的下连交换节点。
步骤c.其他中间交换节点的下连交换节点收到所述聚合可用带宽后，将其作为其接收该信息的上端口到上述源中间交换节点的下连交换节点的聚合可用带宽。
在第二种执行方式中，可选地，当其他中间交换节点收到来自同级源中间交换节点的携带可用带宽信息的消息后，可以仅计算其到所述源中间交换节点的下连交换节点的聚合路径可用带宽。然后中间交换节点将计算得到的聚合可用带宽下发给其下连交换节点，由下连交换节点根据接收到的聚合可用带宽信息和其接收端口的所连链路的上行链路的可用带宽，计算其上端口到上述源中间交换节点的各下连交换节点的聚合可用带宽。计算方法是取各段串联路径的可用带宽的最小值。第二种执行方式的基本步骤如下：
步骤a.各中间交换节点(为了便于描述，将该中间交换节点称为“源中间交换节点”)向其他同级可达的中间交换节点发送携带其各下端口到其下连交换节点的链路的下行可用带宽信息的消息。
步骤b.其他中间交换节点收到上述源中间交换节点的消息后，根据消息中的可用带宽信息、其到源中间交换节点的聚合可用带宽信息，计算其到源中间交换节点的各下连交换节点的聚合可用带宽。计算方法是取各段串联路径的可用带宽的最小值。然后，将计算得到的聚合可用带宽发送给所有下连交换节点。
步骤c.下连交换节点收到所述聚合可用带宽后，根据接收到的聚合可用带宽和其接收端口所连链路的上行链路的可用带宽，计算其上端口到其他同级交换节点的聚合可用带宽。将计算结果作为其接收该信息的上端口到其他同级交换节点的聚合可用带宽。计算方法是取各段串联路径的可用带宽的最小值。
在第三种执行方式中，可选地，当其他中间交换节点收到来自同级源中间交换节点的携带可用带宽信息的消息后，可以将该消息中的带宽信息、其到所述源中间交换节点的聚合带宽信息，都发送给其下连交换节点。由下连交换节点根据收到的带宽信息，以及其接收所述带宽信息的上端口所连链路的上行链路的可用带宽，计算其上端口到所述源中间交换节点的下连交换节点的聚合可用带宽信息。第三种执行方式的基本步骤如下：
步骤a.各中间交换节点(为了便于描述，将该中间交换节点称为“源中间交换节点”)向其他同级可达的中间交换节点发送携带其各下端口到其下连交换节点的链路的下行可用带宽信息的消息。
步骤b.其他中间交换节点收到上述源中间交换节点的消息后，将该消息中的带宽信息(即所述源中间交换节点的各下端口到所连下连交换节点的链路的下行可用带宽信息)、其到所述源中间节点的聚合可用带宽信息，都发送给其下连交换节点。
步骤c.下连交换节点收到其上连中间交换节点发送的带宽信息后，根据其上连中间交换节点到所述源中间交换节点的聚合可用带宽信息，所述源中间交换节点的各下端口到所连下连交换节点的链路的下行可用带宽信息，和其接收带宽信息的上端口所连链路的上行链路的可用带宽，计算其上端口到所述源中间交换节点的各下连交换节点的聚合可用带宽。计算方法是取各段串联路径的可用带宽的最小值。然后，将计算结果作为所述上端口到所述源中间交换节点的下连交换节点的聚合可用带宽。
另外，通过上述三种执行方式的任何一种，下连交换节点可以获得其各上端口到其他同级交换节点的聚合可用带宽。下连交换节点将从不同上端口到其他某一同级交换节点的聚合可用带宽求和，可以得到其到所述其它同级交换节点间聚合可用带宽。可选地，如果下连交换节点已经是边缘交换节点，可以不计算起其到其他同级交换节 点的聚合可用带宽。
下面以图12为例对第一种执行方式进行具体说明。中间交换节点A2.1向交换节点A1.1和A3.1发送携带其下连链路L<A2.1,T2.1>的下行可用带宽信息(值4)和下连链路L<A2.1,T2.2>的下行可用带宽信息(值3)的消息。当中间交换节点A1.1收到所述消息后，计算其各下连交换节点的上端口到A2.1的各下连交换节点的聚合可用带宽。以计算交换节点T1.2的端口P3到交换节点T2.2的聚合可用带宽为例对该计算方法进行说明：取“消息中链路L<A2.1,T2.2>的下行可用带宽(值3)”、“A1.1的路径信息表中A1.1到A2.1的聚合可用带宽(值4)”和“链路L<A1.1,T1.2>的上行可用带宽(值2)”的最小值(值2)作为交换节点T1.2的上端口P3到交换节点T2.2的聚合可用带宽，即取各段路径的可用带宽的最小值作为整条路径的可用带宽。按此方法，A1.1能够计算得到其各下连的交换节点(T1.1和T1.2)的上端口P3到A2.1的各下连交换节点(T2.1和T2.2)的聚合可用带宽。然后A1.1将计算得到的各路径的聚合可用带宽信息发送给路径的源交换节点(A1.1的下连交换节点)。比如，A1.1将T1.2到T2.1的聚合可用带宽(值2)和T1.2到T2.2的聚合可用带宽(值2)发送给T1.2。最终，交换节点T1.1、T1.2、T3.1和T3.2可获得由它们的上端口P3到交换节点T2.1和T2.2的聚合可用带宽，如图12中表3、表4、表5和表6所示。交换节点T*.*将其各上端口到某一交换节点的聚合可用带宽相加，可以进一步计算得到交换节点间的聚合路径的聚合可用带宽。
第二种执行方式和第三中执行方式的执行与第一种执行方式类似，仅计算聚合可用带宽的交换节点有所差异，不再举例赘述。
在本发明实施例中，CLOS交换网络中的第一交换节点可以计算与其连接的至少一个下一级节点的第一目标端口到另一下一级节点之间的路径的聚合可用带宽，以便于下一级交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。提高了测量网络路径的效率。
另外，需要说明的是，如图13所示，现有技术还存在不满足标准CLOS交换网络定义的多级CLOS交换网络的变种。例如，在图13中，根交换节点并非只与每个区块(pod)中的1个交换节点的1个端口相连。但是如图14所示，图14是对图13的多级CLOS交换网络重新划分后获得的标准多级CLOS交换网络。通过将其端口进行划分，我们总能将其转化为逻辑上的标准多级CLOS网络。所以，本发明实施例同样能适用于CLOS交换网络的变种交换网络。
图15示出了本发明实施例的装置1500的示意性框图。装置1500可以是交换机，或者装置1500也可以是具有交换机功能的实体模块。装置1500可以执行图1至图14的方法中由第一交换节点执行的各步骤。装置1500包括：通信模块1510和处理模块1520，
所述处理模块1520用于通过所述通信模块1510获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，所述第一交换节点和所述第二交换节点为所述CLOS交换网络中的同级交换节点；
所述处理模块1520还用于通过所述通信模块1510获取第三交换节点的第一目标端口到所述第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；
所述处理模块1520还用于通过所述通信模块1510获取所述第二交换节点到第四交换节点的路径的第二可用带宽，所述第四交换节点是与所述第二交换节点连接的下一级交换 节点；
所述处理模块1520还用于确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一聚合可用带宽、所述第一可用带宽以及所述第二可用带宽中的最小可用带宽。
在本发明实施例中，CLOS交换网络中的第一交换节点可以通过上述方式计算与其连接的至少一个下一级交换节点的第一目标端口到另一下一级交换节点之间的路径的聚合可用带宽，以便于下一级交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。该方法减少了测量网络路径的复杂度，提高了测量网络路径的效率。
图16示出了本发明实施例的装置1600的示意性框图。装置1600可以是交换机，或者装置1600也可以是具有交换机功能的实体模块。装置1600可以执行图1至图14的方法中由第三交换节点执行的各步骤。装置1600包括：通信模块1610和处理模块1620，
所述处理模块1610用于通过所述通信模块1620获取所述第三交换节点的第一目标端口到第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；
所述处理模块1610还用于通过所述通信模块1620获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，所述第四交换节点与所述第三交换节点为同级交换节点；
所述处理模块1610还用于确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一可用带宽和所述第三聚合可用带宽中较小的可用带宽。
在本发明实施例中，CLOS交换网络中的第三交换节点可以通过上述方式计算其第一目标端口到第四交换节点之间的聚合可用带宽。以便于第三交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。提高了测量网络路径的效率。
图17示出了本发明实施例的装置1700的示意性框图。装置1700可以是交换机，或者装置1700也可以是具有交换机功能的实体模块。装置1700可以执行图1至图14的方法中由第一交换节点执行的各步骤。装置1700包括：
存储器1710，用于存储程序；
通信接口1720，用于和其他设备进行通信；
处理器1730，用于执行存储器1710中的程序，当所述程序被执行时，所述处理器1730用于通过所述通信接口1710获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，所述第一交换节点和所述第二交换节点为所述CLOS交换网络中的同级交换节点；以及通过所述通信接口1710获取第三交换节点的第一目标端口到所述第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；以及通过所述通信接口1710获取所述第二交换节点到第四交换节点的路径的第二可用带宽，所述第四交换节点是与所述第二交换节点连接的下一级交换节点；以及确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一聚合可用带宽、所述第一可用带宽以及所述第二可用带宽中的最小可用带宽。
在本发明实施例中，CLOS交换网络中的第一交换节点可以通过上述方式计算与 其连接的至少一个下一级交换节点的第一目标端口到另一下一级交换节点之间的路径的聚合可用带宽，以便于下一级交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。该方法减少了测量网络路径的复杂度，提高了测量网络路径的效率。
图18示出了本发明实施例的装置1800的示意性框图。装置1800可以是交换机，或者装置1800也可以是具有交换机功能的实体模块。装置1800可以执行图1至图14的方法中由第三交换节点执行的各步骤。装置1800包括：
存储器1810，用于存储程序；
通信接口1820，用于和其他设备进行通信；
处理器1830，用于执行存储器1810存储的程序，当所述程序被执行时，所述处理器1830用于通过所述通信接口1820获取所述第三交换节点的第一目标端口到第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；以及通过所述通信接口1820获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，所述第四交换节点与所述第三交换节点为同级交换节点；以及确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一可用带宽和所述第三聚合可用带宽中较小的可用带宽。
在本发明实施例中，CLOS交换网络中的第三交换节点可以通过上述方式计算其第一目标端口到第四交换节点之间的聚合可用带宽。以便于第三交换节点根据各端口对应的聚合可用带宽，在各端口之间动态调度流量，实现负载均衡。提高了测量网络路径的效率。
另外，本文中术语“系统”和“网络”在本文中常被可互换使用。本文中术语“和/或”，仅仅是一种描述关联对象的关联关系，表示可以存在三种关系，例如，A和/或B，可以表示：单独存在A，同时存在A和B，单独存在B这三种情况。另外，本文中字符“/”，一般表示前后关联对象是一种“或”的关系。
应理解，在本发明实施例中，“与A相应的B”表示B与A相关联，根据A可以确定B。但还应理解，根据A确定B并不意味着仅仅根据A确定B，还可以根据A和/或其他信息确定B。
应理解，在本发明的各种实施例中，上述各过程的序号的大小并不意味着执行顺序的先后，各过程的执行顺序应以其功能和内在逻辑确定，而不应对本发明实施例的实施过程构成任何限定。
可以理解，本发明实施例中的处理器可以是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法实施例的各步骤可以通过处理器中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处理器可以是通用处理器、数字信号处理器(Digital Signal Processor，DSP)、专用集成电路(Application Specific Integrated Circuit，ASIC)、现成可编程门阵列(Field Programmable Gate Array，FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本发明实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本发明实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器，处理器读取存储器中的信息，结合其硬件完成上述方法 的步骤。
可以理解，本发明实施例中的存储器可以是易失性存储器或非易失性存储器，或可包括易失性和非易失性存储器两者。其中，非易失性存储器可以是只读存储器(Read-Only Memory，ROM)、可编程只读存储器(Programmable ROM，PROM)、可擦除可编程只读存储器(Erasable PROM，EPROM)、电可擦除可编程只读存储器(Electrically EPROM，EEPROM)或闪存。易失性存储器可以是随机存取存储器(Random Access Memory，RAM)，其用作外部高速缓存。通过示例性但不是限制性说明，许多形式的RAM可用，例如静态随机存取存储器(Static RAM，SRAM)、动态随机存取存储器(Dynamic RAM，DRAM)、同步动态随机存取存储器(Synchronous DRAM，SDRAM)、双倍数据速率同步动态随机存取存储器(Double Data Rate SDRAM，DDR SDRAM)、增强型同步动态随机存取存储器(Enhanced SDRAM，ESDRAM)、同步连接动态随机存取存储器(Synchlink DRAM，SLDRAM)和直接内存总线随机存取存储器(Direct Rambus RAM，DR RAM)。应注意，本文描述的系统和方法的存储器旨在包括但不限于这些和任意其他适合类型的存储器。
本领域普通技术人员可以意识到，结合本文中所公开的实施例描述的各示例的单元及算法步骤，能够以电子硬件、或者计算机软件和电子硬件的结合来实现。这些功能究竟以硬件还是软件方式来执行，取决于技术方案的特定应用和设计约束条件。专业技术人员可以对每个特定的应用来使用不同方法来实现所描述的功能，但是这种实现不应认为超出本发明的范围。
所属领域的技术人员可以清楚地了解到，为描述的方便和简洁，上述描述的系统、装置和单元的具体工作过程，可以参考前述方法实施例中的对应过程，在此不再赘述。
在本申请所提供的几个实施例中，应该理解到，所揭露的系统、装置和方法，可以通过其他的方式实现。例如，以上所描述的装置实施例仅仅是示意性的，例如，所述单元的划分，仅仅为一种逻辑功能划分，实际实现时可以有另外的划分方式，例如多个单元或组件可以结合或者可以集成到另一个系统，或一些特征可以忽略，或不执行。另一点，所显示或讨论的相互之间的耦合或直接耦合或通信连接可以是通过一些接口，装置或单元的间接耦合或通信连接，可以是电性，机械或其他的形式。
所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部单元来实现本实施例方案的目的。
另外，在本发明各个实施例中的各功能单元可以集成在一个处理单元中，也可以是各个单元单独物理存在，也可以两个或两个以上单元集成在一个单元中。
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本发明的技术方案本质上或者说对现有技术做出贡献的部分或者该技术方案的部分可以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中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本发明各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等各种可以存储程序代码的介质。
以上所述，仅为本发明的具体实施方式，但本发明的保护范围并不局限于此，任何熟悉本技术领域的技术人员在本发明揭露的技术范围内，可轻易想到变化或替换，都应涵盖在本发明的保护范围之内。因此，本发明的保护范围应以权利要求的保护范围为准。

Claims (20)

  1. 一种用于测量网络路径的方法，其特征在于，包括：
    CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，所述第一交换节点和所述第二交换节点为所述CLOS交换网络中的同级交换节点；
    所述第一交换节点获取第三交换节点的第一目标端口到所述第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；
    所述第一交换节点获取所述第二交换节点到第四交换节点的路径的第二可用带宽，所述第四交换节点是与所述第二交换节点连接的下一级交换节点；
    所述第一交换节点确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一聚合可用带宽、所述第一可用带宽以及所述第二可用带宽中的最小可用带宽。
  2. 如权利要求1所述的方法，其特征在于，所述CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，包括：
    所述第一交换节点分别从多个第五交换节点接收第一信息，所述第一信息指示所述多个第五交换节点中的每个第五交换节点对应的第一路径的聚合可用带宽，其中，所述多个第五交换节点是与所述第一交换节点连接的上一级交换节点，所述第一路径是经过所述每个第五交换节点的所述第一交换节点到所述第二交换节点的路径；
    所述第一交换节点根据所述第一信息，将所述多个第五交换节点对应的第一路径的聚合可用带宽之和确定为所述第一聚合可用带宽。
  3. 如权利要求1所述的方法，其特征在于，所述第一交换节点和所述第二交换节点是与所述CLOS交换网络中的多个第六交换节点连接的下一级交换节点，
    所述CLOS交换网络中的第一交换节点获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，包括：
    所述第一交换节点获取所述第一交换节点到所述多个第六交换节点中的每个第六交换节点的路径的上行可用带宽；
    所述第一交换节点从所述每个第六交换节点接收第二信息，所述第二信息用于指示所述每个第六交换节点到所述第二交换节点的下行可用带宽；
    所述第一交换节点确定所述上行可用带宽和所述下行可用带宽中较小的可用带宽为所述每个第六交换节点对应的可用带宽；
    将所述多个第六交换节点对应的可用带宽之和确定为所述第一聚合可用带宽。
  4. 如权利要求1至3中任一项所述的方法，其特征在于，所述第一交换节点获取所述第二交换节点到第四交换节点的路径的第二可用带宽，包括：
    所述第一交换节点接收所述第二交换节点发送的第三信息，所述第三信息用于指示所述第二交换节点到第四交换节点的路径的可用带宽是所述第二可用带宽。
  5. 如权利要求1至4中任一项所述的方法，其特征在于，所述方法还包括：所述第一交换节点向所述第三交换节点发送第四信息，所述第四信息用于指示所述第三交换节点的第一目标端口到所述第四交换节点之间的可用带宽是所述第二聚合可用带宽。
  6. 一种用于测量网络路径的方法，其特征在于，包括：
    CLOS交换网络中的第三交换节点获取所述第三交换节点的第一目标端口到第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；
    所述第三交换节点获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，所述第四交换节点与所述第三交换节点为同级交换节点；
    所述第三交换节点确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一可用带宽和所述第三聚合可用带宽中较小的可用带宽。
  7. 如权利要求6所述的方法，其特征在于，所述第三聚合可用带宽是第一聚合可用带宽和第二可用带宽中的较小的可用带宽，所述第一聚合可用带宽是所述第一交换节点到第二交换节点的路径的聚合可用带宽，所述第二可用带宽是所述第二交换节点到所述第四交换节点的路径的第二可用带宽，所述第二交换节点是与所述第四交换节点连接的上一级交换节点。
  8. 如权利要求6或7所述的方法，其特征在于，所述第三交换节点获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，包括：
    所述第三交换节点从所述第一交换节点接收第一指示信息，所述第一指示信息用于指示所述第一交换节点到所述第四交换节点的路径的聚合可用带宽为所述第三聚合可用带宽。
  9. 如权利要求6或7所述的方法，其特征在于，所述第三交换节点获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，包括：
    所述第三交换节点接收所述第一交换节点发送的第二指示信息，所述第二指示信息用于指示所述第一交换节点到第二交换节点的路径的第一聚合可用带宽；
    所述第三交换节点接收所述第二交换节点发送的第三指示信息，所述第三指示信息用于指示所述第二交换节点到所述第四交换节点的路径的第二可用带宽；
    所述第三交换节点确定所述第三聚合可用带宽，所述第三聚合可用带宽是第一聚合可用带宽和第二可用带宽中的较小的可用带宽。
  10. 如权利要求6至9中任一项所述的方法，其特征在于，所述方法还包括：所述第三交换节点确定所述第三交换节点到所述第四交换节点的第四聚合可用带宽，所述第四聚合可用带宽为所述第三交换节点的各端口分别到所述第四交换节点的路径的聚合可用带宽之和。
  11. 一种用于测量网络路径的装置，其特征在于，所述装置为CLOS交换网络中的第一交换节点，包括：通信模块和处理模块，
    所述处理模块用于通过所述通信模块获取所述第一交换节点到第二交换节点的路径的第一聚合可用带宽，所述第一交换节点和所述第二交换节点为所述CLOS交换网络中的同级交换节点；
    所述处理模块还用于通过所述通信模块获取第三交换节点的第一目标端口到所述第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；
    所述处理模块还用于通过所述通信模块获取所述第二交换节点到第四交换节点的路径的第二可用带宽，所述第四交换节点是与所述第二交换节点连接的下一级交换节点；
    所述处理模块还用于确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一聚合可用带宽、所述第一可用带宽以及所述第二可用带宽中的最小可用带宽。
  12. 如权利要求11所述的装置，其特征在于，所述处理模块具体用于通过所述通信模块分别从多个第五交换节点接收第一信息，所述第一信息指示所述多个第五交换节点中的每个第五交换节点对应的第一路径的聚合可用带宽，其中，所述多个第五交换节点是与所述第一交换节点连接的上一级交换节点，所述第一路径是经过所述每个第五交换节点的所述第一交换节点到所述第二交换节点的路径；以及根据所述第一信息，将所述多个第五交换节点对应的第一路径的聚合可用带宽之和确定为所述第一聚合可用带宽。
  13. 如权利要求11所述的装置，其特征在于，所述第一交换节点和所述第二交换节点是与所述CLOS交换网络中的多个第六交换节点连接的下一级交换节点，所述处理模块具体用于通过所述通信模块获取所述第一交换节点到所述多个第六交换节点中的每个第六交换节点的路径的上行可用带宽；以及通过所述通信模块从所述每个第六交换节点接收第二信息，所述第二信息用于指示所述每个第六交换节点到所述第二交换节点的下行可用带宽；以及确定所述上行可用带宽和所述下行可用带宽中较小的可用带宽为所述每个第六交换节点对应的可用带宽；以及将所述多个第六交换节点对应的可用带宽之和确定为所述第一聚合可用带宽。
  14. 如权利要求11至13中任一项所述的装置，其特征在于，所述处理模块具体用于通过所述通信模块接收所述第二交换节点发送的第三信息，所述第三信息用于指示所述第二交换节点到第四交换节点的路径的可用带宽是所述第二可用带宽。
  15. 如权利要求11至14中任一项所述的装置，其特征在于，所述处理模块还用于通过所述通信模块向所述第三交换节点发送第四信息，所述第四信息用于指示所述第三交换节点的第一目标端口到所述第四交换节点之间的可用带宽是所述第二聚合可用带宽。
  16. 一种用于测量网络路径的装置，其特征在于，所述装置为CLOS交换网络中的第三交换节点，包括处理模块和通信模块：
    所述处理模块用于通过所述通信模块获取所述第三交换节点的第一目标端口到第一交换节点的路径的第一可用带宽，所述第三交换节点是与所述第一交换节点连接的下一级交换节点；
    所述处理模块还用于通过所述通信模块获取所述第一交换节点到第四交换节点的路径的第三聚合可用带宽，所述第四交换节点与所述第三交换节点为同级交换节点；
    所述处理模块还用于确定所述第三交换节点的第一目标端口到所述第四交换节点的路径的第二聚合可用带宽，所述第二聚合可用带宽为所述第一可用带宽和所述第三聚合可用带宽中较小的可用带宽。
  17. 如权利要求16所述的装置，其特征在于，所述第三聚合可用带宽是第一聚合可用带宽和第二可用带宽中的较小的可用带宽，所述第一聚合可用带宽是所述第一交换节点到第二交换节点的路径的聚合可用带宽，所述第二可用带宽是所述第二交换节点到所述第四交换节点的路径的第二可用带宽，所述第二交换节点是与所述第四交换节点连接的上一级交换节点。
  18. 如权利要求16或17所述的装置，其特征在于，所述处理模块具体用于通过所述通信模块从所述第一交换节点接收第一指示信息，所述第一指示信息用于指示所述第一交 换节点到所述第四交换节点的路径的聚合可用带宽为所述第三聚合可用带宽。
  19. 如权利要求16或17所述的装置，其特征在于，所述处理模块具体用于通过所述通信模块接收所述第一交换节点发送的第二指示信息，所述第二指示信息用于指示所述第一交换节点到第二交换节点的路径的第一聚合可用带宽；以及通过所述通信模块接收所述第二交换节点发送的第三指示信息，所述第三指示信息用于指示所述第二交换节点到所述第四交换节点的路径的第二可用带宽；以及确定所述第三聚合可用带宽，所述第三聚合可用带宽是第一聚合可用带宽和第二可用带宽中的较小的可用带宽。
  20. 如权利要求16至19中任一项所述的装置，其特征在于，所述处理模块还用于确定所述第三交换节点到所述第四交换节点的第四聚合可用带宽，所述第四聚合可用带宽为所述第三交换节点的各端口分别到所述第四交换节点的路径的聚合可用带宽之和。
PCT/CN2017/106557 2016-10-17 2017-10-17 用于测量网络路径的方法、装置和系统 WO2018072690A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP17862716.2A EP3515018B1 (en) 2016-10-17 2017-10-17 Method, apparatus and system for measuring network path
US16/385,954 US11082358B2 (en) 2016-10-17 2019-04-16 Network path measurement method, apparatus, and system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201610902770.4A CN107959642B (zh) 2016-10-17 2016-10-17 用于测量网络路径的方法、装置和系统
CN201610902770.4 2016-10-17

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US16/385,954 Continuation US11082358B2 (en) 2016-10-17 2019-04-16 Network path measurement method, apparatus, and system

Publications (1)

Publication Number Publication Date
WO2018072690A1 true WO2018072690A1 (zh) 2018-04-26

Family

ID=61954516

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/106557 WO2018072690A1 (zh) 2016-10-17 2017-10-17 用于测量网络路径的方法、装置和系统

Country Status (4)

Country Link
US (1) US11082358B2 (zh)
EP (1) EP3515018B1 (zh)
CN (1) CN107959642B (zh)
WO (1) WO2018072690A1 (zh)

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107592269B (zh) * 2016-07-06 2020-03-31 华为技术有限公司 传输路径的负载信息的方法和网络节点
US11146505B2 (en) * 2018-01-11 2021-10-12 Intel Corporation Low latency compact Clos network controller
CN111447103B (zh) * 2020-03-09 2022-01-28 杭州海康威视系统技术有限公司 虚拟设备的管理系统及方法、电子设备及介质
CN114896192B (zh) * 2022-06-09 2024-04-09 苏州华兴源创科技股份有限公司 接口的匹配方法、装置、计算机设备和存储介质

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020146003A1 (en) * 2001-03-19 2002-10-10 Kam Anthony Chi-Kong Traffic spreading to reduce blocking in a groomed CLOS communication switch
CN102835081A (zh) * 2012-05-21 2012-12-19 华为技术有限公司 基于三级互联交换网络的调度方法、装置及系统
CN105706400A (zh) * 2013-11-05 2016-06-22 思科技术公司 网络结构覆盖
CN105763469A (zh) * 2016-04-07 2016-07-13 烽火通信科技股份有限公司 三级Clos网络架构中链路拥塞检测及带宽控制的方法与系统

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
SE0004303D0 (sv) * 2000-11-23 2000-11-23 Net Insight Ab Switching apparatus
US7586909B1 (en) * 2002-03-06 2009-09-08 Agere Systems Inc. Striping algorithm for switching fabric
US20070121499A1 (en) * 2005-11-28 2007-05-31 Subhasis Pal Method of and system for physically distributed, logically shared, and data slice-synchronized shared memory switching
US20070253334A1 (en) * 2006-04-26 2007-11-01 Chetan Mehta Switch routing algorithm for improved congestion control & load balancing
US9106564B2 (en) 2011-03-31 2015-08-11 Amazon Technologies, Inc. Incremental high radix network scaling
CN104380679A (zh) * 2012-05-01 2015-02-25 惠普发展公司，有限责任合伙企业 可配置的clos网络
US8867560B2 (en) * 2012-07-30 2014-10-21 Cisco Technology, Inc. Managing crossbar oversubscription
US9166886B1 (en) * 2013-06-19 2015-10-20 Google Inc. Systems and methods for determining physical network topology
US10291503B2 (en) * 2013-09-26 2019-05-14 Taiwan Semiconductor Manufacturing Co., Ltd. File block placement in a distributed network
US9538264B2 (en) * 2014-08-07 2017-01-03 Ciena Corporation ODUflex resizing systems and methods
US9716669B2 (en) * 2014-12-04 2017-07-25 Juniper Networks, Inc. Multi-chassis switch having a modular center stage chassis
US9813356B1 (en) * 2016-02-11 2017-11-07 Amazon Technologies, Inc. Calculating bandwidth information in multi-stage networks

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20020146003A1 (en) * 2001-03-19 2002-10-10 Kam Anthony Chi-Kong Traffic spreading to reduce blocking in a groomed CLOS communication switch
CN102835081A (zh) * 2012-05-21 2012-12-19 华为技术有限公司 基于三级互联交换网络的调度方法、装置及系统
CN105706400A (zh) * 2013-11-05 2016-06-22 思科技术公司 网络结构覆盖
CN105763469A (zh) * 2016-04-07 2016-07-13 烽火通信科技股份有限公司 三级Clos网络架构中链路拥塞检测及带宽控制的方法与系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3515018A4 *

Also Published As

Publication number Publication date
EP3515018B1 (en) 2023-04-05
CN107959642B (zh) 2020-08-07
EP3515018A1 (en) 2019-07-24
US11082358B2 (en) 2021-08-03
US20190245803A1 (en) 2019-08-08
CN107959642A (zh) 2018-04-24
EP3515018A4 (en) 2019-07-24

Similar Documents

Publication Publication Date Title
JP6093867B2 (ja) インターコネクトにおける不均一なチャネル容量
US10735323B2 (en) Service traffic allocation method and apparatus
WO2018072690A1 (zh) 用于测量网络路径的方法、装置和系统
US10623303B2 (en) Method of routing data and switch in a network
CN112104558B (zh) 区块链分发网络的实现方法、系统、终端及介质
US9600440B2 (en) Network topology of hierarchical ring with recursive shortcuts
CN107659426B (zh) 分配物理资源的方法和网络侧设备
Ansari et al. Performance evaluation of various parameters of Network-on-Chip (NoC) for different topologies
US9529775B2 (en) Network topology of hierarchical ring with gray code and binary code
CN109218190B (zh) 一种传输路径的确定方法及节点
WO2018137361A1 (zh) 数据转发方法及装置
US10637739B2 (en) Network topology system and building method for topologies and routing tables thereof
CN117135059B (zh) 一种网络拓扑结构、构造方法、路由算法、设备及介质
CN116915708A (zh) 路由数据包的方法、处理器及可读存储介质
WO2023093513A1 (zh) 路径感知方法、装置及系统
Kojić et al. New algorithm for packet routing in mobile ad-hoc networks
Cai et al. Deadlock-free adaptive routing based on the repetitive turn model for 3D network-on-chip
Punhani et al. E-XY: an entropy based XY routing algorithm
Liu et al. On the min-max-delay problem: NP-completeness, algorithm, and integrality gap
EP3471351B1 (en) Method and device for acquiring path information about data packet
CN105900383A (zh) 通信系统、控制节点和通信方法
Grewal et al. Load Balanced Routing Approach using the Routing Path Status Awareness for Effective Data Propagation
CN116192753A (zh) 一种流量分配方法及装置
Aslam et al. High Performance and Connection-retaining Fault-Tolerant Routing Algorithm for NoCbased many core Systems
Kulkarni et al. On multipath routing algorithm for software defined networks

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: 17862716

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

ENP Entry into the national phase

Ref document number: 2017862716

Country of ref document: EP

Effective date: 20190418