CN116094962A - Network transmission delay measurement method for any two nodes in programmable network - Google Patents

Network transmission delay measurement method for any two nodes in programmable network Download PDF

Info

Publication number
CN116094962A
CN116094962A CN202310082418.0A CN202310082418A CN116094962A CN 116094962 A CN116094962 A CN 116094962A CN 202310082418 A CN202310082418 A CN 202310082418A CN 116094962 A CN116094962 A CN 116094962A
Authority
CN
China
Prior art keywords
node
transmission delay
abnormal
network
flow
Prior art date
Legal status (The legal status 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 status listed.)
Pending
Application number
CN202310082418.0A
Other languages
Chinese (zh)
Inventor
闫林林
葛俊成
卢东辉
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhejiang Lab
Original Assignee
Zhejiang Lab
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 Zhejiang Lab filed Critical Zhejiang Lab
Priority to CN202310082418.0A priority Critical patent/CN116094962A/en
Publication of CN116094962A publication Critical patent/CN116094962A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • 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
    • H04L43/0852—Delays
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/22—Indexing; Data structures therefor; Storage structures
    • G06F16/2228—Indexing structures
    • G06F16/2255—Hash tables
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
    • G06F16/24—Querying
    • G06F16/245—Query processing
    • G06F16/2455—Query execution
    • 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
    • H04L43/02—Capturing of monitoring data
    • H04L43/028—Capturing of monitoring data by filtering
    • 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
    • H04L43/04—Processing captured monitoring data, e.g. for logfile generation
    • 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
    • H04L43/10—Active monitoring, e.g. heartbeat, ping or trace-route
    • H04L43/106—Active monitoring, e.g. heartbeat, ping or trace-route using time related information in packets, e.g. by adding timestamps
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/50—Reducing energy consumption in communication networks in wire-line communication networks, e.g. low power modes or reduced link rate

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Data Mining & Analysis (AREA)
  • General Physics & Mathematics (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Cardiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Environmental & Geological Engineering (AREA)
  • Computational Linguistics (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention discloses a method for measuring network transmission delay of any two nodes in a programmable network, which comprises the steps of firstly adding a time stamp label to a data packet at a node A and obtaining transmission delay d at a node B l And filtering according to a time delay threshold delta, forwarding the abnormal flow to a flow processor, and then in a bloom filter, if the record of the abnormal flow exists, considering the abnormal flow as the abnormal flow and forwarding the abnormal flow to the flow processor, otherwise, forwarding the abnormal flow to a Sketch algorithm module. In-stream processingAnd in the device, the five-tuple and the accumulation time delay of the abnormal flow are accurately recorded by using a hash table for the abnormal flow, so that the controller can conveniently and rapidly locate the generated abnormal flow. And adding the accumulated sum of the transmission delays in the Sketch algorithm module, so that the average transmission delay can be measured by combining the number of the data packets and can be used as the network transmission delay measurement. The invention reduces the resource consumption, and can rapidly locate abnormal flow through the hash table, and further improve the measurement accuracy of CM sktch after filtering abnormal values.

Description

Network transmission delay measurement method for any two nodes in programmable network
Technical Field
The invention belongs to the technical field of network management, and particularly relates to a network transmission delay measurement method for any two nodes in a programmable network.
Background
As the internet scale continues to expand, its performance becomes more and more complex, and the complexity of network management increases. In actual network operation, as large as an internet network center and a service provider need to monitor the current operation state of the network through a network performance measurement system, as small as a single user and an application program need to detect the operation state of the network through the network performance measurement system and evaluate the service quality provided by the network, so that research on network performance measurement becomes more and more important along with the wider application of the internet.
Currently, basic network performance evaluation indexes mainly comprise transmission delay, throughput (bandwidth), packet loss rate and the like of a network. Among these, the transmission delay of the network is the most important and most commonly used measurement parameter. In order to ensure the normal operation of the network and meet the transmission performance requirements of network applications, a network manager needs to use a specific measurement means and a measurement system to measure and analyze the transmission delay of the network, know the operation state inside the network timely, accurately and comprehensively, and implement a network management and control strategy in a targeted manner. The transmission delay of the network can provide basis for network fault diagnosis, help detect abnormal traffic flows and locate node faults in the network. Generally, the finer the granularity of the measured data of the network transmission delay, the higher the accuracy, and the better the network management can achieve the performance.
The measurement scheme in the traditional IP network is generally difficult to monitor network transmission delay in real time, measurement equipment is difficult to deploy, network resource consumption is overlarge, measurement granularity is insufficient, and the like. Under the limitation of the traditional network control and measurement capability, the network transmission delay measurement provides low data accuracy and insufficient fine granularity, and even in many occasions, the network manager cannot be provided with the required information, and the following problems mainly exist: (1) Control logic of network equipment in a traditional network is integrated with forwarding logic, and the processing of data packets by a switch needs to follow certain specific protocols, so that flexible control capability is lacked. (2) In the traditional network, the measurement capability of the network equipment is low, and only some coarse-grained state information can be measured, so that the measurement error is larger. (3) Some measurement tasks in the network need cooperation among a plurality of network devices, but the traditional network adopts a distributed control mode, so that the cooperation and management overhead among network nodes is large, the feasibility of many measurement tasks only stays in theory, and the measurement tasks cannot be deployed in an actual network at all.
Disclosure of Invention
The invention aims to overcome the defects of the prior art, and provides a network transmission delay measurement method for any two nodes in a programmable network.
In order to achieve the above object, the method for measuring network transmission delay of any two nodes in a programmable network according to the present invention is characterized by comprising the following steps:
(1) Packet timestamp label addition and abnormal time delay stream judgment
Service request f initiated to network for a content 1 The network transmission delay from node A to node B needs to be measured, a time stamp label is added to the data packet at node A, and the content of the time stamp label is the entry time stamp t of the data packet reaching the switch A 1 When a data packet is forwarded to the node B, the node B obtains the current timestamp t 2 Then analyzing the timestamp label in the data packet to obtain an entry timestamp t1, and then requesting the service f 1 Transmission delay d between node a and node B l ＝t 2 -t 1 ；
In node B, the transmission delay d is calculated l After that, the transmission delay d l Compared with a time delay threshold delta, the transmission time delay d l Service request f greater than delay threshold delta 1 To transmit an abnormal stream, which is a stream with abnormal performance, five-tuple of the abnormal stream is encoded and used with clothThe bloom filter records, and at the same time, the abnormal flow is forwarded to the flow processor; for transmission delay d l Service request f not greater than delay threshold delta 1 Sending a bloom filter to make further decisions, wherein the bloom filter is deployed in the node B;
(2) The bloom filter further separates the abnormal flow
In bloom filter, for service request f 1 The five-tuple, namely the source IP, the destination IP, the source port, the destination port and the protocol type, are encoded by adopting the same encoding method in the step (1), and compared with the five-tuple encoding of the abnormal flow recorded in the bloom filter, if the five-tuple is the same, the five-tuple is considered as the abnormal flow and is forwarded to a flow processor, if the five-tuple is not the same, the five-tuple is the normal flow and is forwarded to a Sketch algorithm module, wherein the Sketch algorithm module is deployed in a node B;
(3) The stream processor records the abnormal stream and reports the abnormal stream to the controller for positioning
The abnormal flow judged by the time delay threshold delta and the abnormal flow separated by the bloom filter are accurately recorded in the flow processor by using a hash table, the five-tuple and the accumulated time delay of the abnormal flow are reported to the controller according to the measurement period, and the controller can rapidly locate the generated abnormal flow through the hash table;
(4) Network transmission delay measurement by using Sketch algorithm
In the sktch algorithm module, each counter contains two parts: the count value count of 32 bits is used for counting the number of data packets and the count value sumdelay of 32 bits is used for storing the accumulated sum of transmission delay;
when the transmission delay d is calculated l The node B then passes the five-tuple of the data packet through the hash function h 1 To h d D times of hashing are carried out, a counter corresponding to each time of hashing is found to update, and the updating rule is as follows: count=count+1, sumdelay=sumdelay+d l ；
The node B reports the count value count and the count value sumdelay of the counter to the controller according to the measurement period described in the step (3), and the controller inquires the service through the count value count and the count value sumdelayStream transmission delay, inquiry service request f 1 When the transmission delay information of (1) is transmitted, the hash functions h1 to h are firstly used for d For service request f 1 Performing hash for d times to find corresponding d counters, and service request f 1 Query result count of number of data packets s For the minimum value in the count value count of d counter counters, the service request f 1 Query result sumdelay of data packet transmission delay value sum s Calculating the minimum value in the sumdelay value for d counter counters;
within one period, service request f 1 The average transmission delay between node a and node B is:
Figure BDA0004067835620000031
taking the average transmission delay as the network transmission delay from the node A to the node B;
(5) Emptying bloom filter, counter and hash table
After each measurement cycle is completed, the bloom filter, counter in the sktech algorithm module, and hash table in the stream processor are emptied, ready for the next measurement cycle.
The invention aims at realizing the following steps:
the invention relates to a network transmission delay measuring method of any two nodes in a programmable network, which comprises the steps of firstly adding a time stamp label to a data packet at a node A, and obtaining a transmission delay d at the node B when the data packet is forwarded to the node B l And filtering according to a time delay threshold delta, forwarding the abnormal flow to a flow processor, sending the abnormal flow to a bloom filter which is not identified for further judgment, and then in a bloom filter, if the record of the abnormal flow exists, considering the abnormal flow as the abnormal flow and forwarding the abnormal flow to the flow processor, otherwise, forwarding the abnormal flow to a Sketch algorithm module. In the stream processor, the five-tuple and the accumulation delay of the abnormal stream are accurately recorded by a hash table for the abnormal stream, so that the controller can conveniently and rapidly locate the generated abnormal stream. The transmission delay summation is added in the Sketch algorithm module, so that the number of the data packets can be measuredThe average transmission delay is taken out and used as the network transmission delay measurement. The invention adopts the data packet to add the timestamp label, thus when the measurement algorithm is deployed, only the node B is required to provide a small amount of high-speed storage resources, the resource consumption is reduced, and meanwhile, the invention separates the flow with abnormal network performance again through the bloom filter, on one hand, the abnormal flow can be rapidly positioned through the hash table, and on the other hand, after the abnormal value is filtered, the measurement accuracy of CM (common channel) Sketch can be further improved.
Drawings
FIG. 1 is a schematic diagram of a method for measuring network transmission delay of any two nodes in a programmable network according to the present invention;
FIG. 2 is a flow chart of one embodiment of a method for measuring network transmission delay of any two nodes in a programmable network according to the present invention;
fig. 3 is a schematic diagram of a data structure in the present invention.
Detailed Description
The following description of the embodiments of the invention is presented in conjunction with the accompanying drawings to provide a better understanding of the invention to those skilled in the art. It is to be expressly noted that in the description below, detailed descriptions of known functions and designs are omitted here as perhaps obscuring the present invention.
As described in the background art, the conventional network transmission delay measurement algorithm has a certain limitation, and cannot be deployed in an actual network at all. The programmable network presents the data packet processing logic and the network control logic directly to the user. The data plane has the capability of customizing the data packet processing flow, and the control plane provides centralized, flexible and quick control capability. The method mainly utilizes a network application programming interface which provides a standard in a network node to provide an open network control mechanism for a user, so that the flexible processing capacity is convenient for the implementation of a measurement task, and the accurate, timely and fine-granularity network transmission delay measurement in a network is possible.
Under programmable network architecture, to achieve routing of traffic flows on any two pathsFine-grained measurement of transmission delay between individual nodes, assuming that a measurement service request f is required 1 The transmission delay from node A to node B on the path is realized by the following simpler modes: respectively deploying hash tables on the node A and the node B, wherein the hash tables can record quintuple of each service flow as key, the time stamp of a data packet reaching the node is taken as value, the switching node can periodically report statistical information to the controller, and the controller can calculate the service request f according to the quintuple and the time stamp recorded by the node A and the node B 1 Transmission delay from node a to node B.
However, the high-speed memory resources of programmable switches are very limited due to cost and physical constraints, and special forwarding chips are required to implement in-band network telemetry, which are expensive. Furthermore, the flexible packet processing capability of the programmable network is implemented by flow tables, the number of which is limited by TCAM (Temary Content Aware Memory, TCAM) size. In the above solution, in order to store five tuples and timestamps, each traffic flow needs to occupy 17 bytes, and there are tens or even millions of traffic flows in each reporting period of the switch, which requires the programmable switch to provide high-speed storage resources of tens of megabits, which is costly.
In order to save the high-speed storage resources of the programmable switch as much as possible, the existing solution provides a Sketch algorithm for us, on one hand, the Sketch algorithm only needs a simple hash function and storage resources, which are easy to realize on the programmable switch, and on the other hand, by compressing the measured information, the Sketch algorithm can save the storage space greatly, which is important for the programmable switch with limited high-speed storage resources. Meanwhile, the programmability of the programmable network data surface provides possibility for further optimization, and the invention can calculate the service flow f according to the current time stamp by adding the time stamp label into the data packet at the node A and analyzing the label by the node B 1 The transmission delay from node A to node B on the path is reduced, so that the node B only needs to provide a small amount of high-speed storage resources when the measurement algorithm is deployed.
The Sketch algorithm, while solving the problem of limited high-speed memory resources of programmable switches, has its drawbacks as well. Firstly, because the measured information is compressed, quintuple information of the service flow cannot be obtained directly through the sktech, and secondly, the problem of measurement accuracy is brought. When the traffic stream encounters problems such as network congestion or node failure during transmission, the transmission delay of the traffic stream may be increased up to tens to hundreds of milliseconds, the original normal traffic stream may have a delay of only a few milliseconds, the normal traffic stream is polluted by the measurement result of the abnormal traffic stream due to the hash collision problem existing in the sktech, the sktech cannot directly acquire quintuple information, thus the measurement result of the normal traffic stream is higher, and the abnormal traffic stream cannot be rapidly and accurately positioned. To solve this problem, the present invention introduces a bloom filter because the main purpose of network transmission delay measurement is to know the network operation state and find out the network performance abnormality for the network manager. Therefore, how to more accurately measure and separate out the abnormal flow of transmission performance in the network is of practical significance. The bloom filter is used for separating the flow with abnormal network performance, codes and records the id of the abnormal flow, and then forwards the abnormal flow to a hash table in a flow processor, wherein the hash table can accurately record the id and measurement information of the abnormal flow. Meanwhile, the Sketch algorithm module records the stream with normal transmission performance and periodically reports the statistical information to the controller. After separating the abnormal flow of network performance through the bloom filter, on one hand, the abnormal flow can be rapidly positioned through the hash table in the flow processor, and on the other hand, after filtering abnormal values, the measurement accuracy of the Sketch algorithm module can be further improved.
The invention uses the characteristics of the programmable network and uses a bloom filter and a Sketch algorithm module to measure the network transmission delay. The method for measuring the network transmission delay for distinguishing abnormal flows under the programmable network is provided by considering the precious storage resources of the programmable switch and the practical significance of the abnormal flows of the transmission delay in knowing the network operation state.
Compared with the traditional network transmission delay measurement method, the invention has the following characteristics:
1. adding a timestamp label to the programmable network;
2. separating the flow with abnormal transmission performance in the network through a bloom filter;
3. using a sktech algorithm to approximately measure network transmission delay;
4. the five-tuple and the accumulated delay are recorded by using a hash table of the stream processor, and the hash table is used for quickly positioning abnormal streams.
Fig. 1 is a schematic diagram of a network transmission delay measurement method of any two nodes in a programmable network according to the present invention.
As shown in fig. 1, the present invention is mainly divided into three parts, and the functions thereof are divided as follows:
1. bloom filter
The stream processor is used for encoding and recording the id of the abnormal performance stream and forwarding the stream which does not pass through the bloom filter to the stream processor for processing;
2. Count-Min Sketch algorithm module
The method is used for recording the transmission delay information of any two points of the non-abnormal flow on each flow routing path.
3. Stream processor
And the five-tuple and the accumulation time delay of the abnormal flow are accurately recorded by using the hash table for the abnormal flow, so that the controller can conveniently and rapidly locate the generated abnormal flow.
Fig. 2 is a flowchart of an embodiment of a method for measuring network transmission delay of any two nodes in a programmable network according to the present invention.
In this embodiment, as shown in fig. 2, the network transmission delay measurement method of any two nodes in the programmable network of the present invention includes the following steps:
step S1: packet timestamp label addition and abnormal time delay stream judgment
Service request f initiated to network for a content 1 The network transmission delay from node A to node B needs to be measured, a time stamp label is added to the data packet at node A, and the content of the time stamp label is the entry time stamp t of the data packet reaching the switch A 1 When a data packet is forwarded to node B, node B obtains the time when it wasFront timestamp t 2 Then analyzing the time stamp label in the data packet to obtain an entry time stamp t 1 Service request f1 At the position of Transmission delay d from node a to node B l ＝t 2 -t 1 。
In node B, as shown in fig. 3, a transmission delay d is calculated l After that, the transmission delay d l Compared with a time delay threshold delta, the transmission time delay d l Service request f greater than delay threshold delta 1 Encoding five-tuple of the abnormal flow and recording by using a bloom filter for transmitting the flow with abnormal performance, namely the abnormal flow, and forwarding the abnormal flow to a flow processor; for transmission delay d l Service request f not greater than delay threshold delta 1 I.e., flows that are not deemed to be non-anomalous, are fed into a bloom filter for further determination, wherein the bloom filter is deployed in the node B.
Step S2: bloom filters further separate abnormal flows
As shown in fig. 3, in the bloom filter, a service request f is made 1 The five-tuple, namely the source IP, the destination IP, the source port, the destination port and the protocol type, are encoded by adopting the same encoding method in the step (1), compared with the five-tuple encoding of the abnormal flow recorded in the bloom filter, if the five-tuple is the same, the five-tuple is considered as the abnormal flow and forwarded to the flow processor, if the five-tuple is not the same, the five-tuple is the normal flow and forwarded to the Sketch algorithm module, wherein the Sketch algorithm module is deployed in the node B.
Step S3: the stream processor records the abnormal stream and reports the abnormal stream to the controller for positioning
The abnormal flow judged by the time delay threshold delta and the abnormal flow separated by the bloom filter are accurately recorded in the flow processor by using a hash table, the five-tuple and the accumulated time delay of the abnormal flow are reported to the controller according to the measurement period, and the controller can rapidly locate the generated abnormal flow through the hash table.
Step S4: network transmission delay measurement by using Sketch algorithm
In the sktch algorithm module, each counter contains two parts: the count value count of 32 bits is used for counting the number of data packets and the count value sumdelay of 32 bits is used for storing the accumulated sum of transmission delay;
as shown in FIG. 3, when the transmission delay d is found l The node B then passes the five-tuple of the data packet through the hash function h 1 To h d D times of hashing are carried out, a counter corresponding to each time of hashing is found to update, and the updating rule is as follows: count=count+1, sumdelay=sumdelay+d l 。
The node B reports the count value count and the count value sumdelay of the counter to the controller according to the measurement period described in the step (3), and the controller inquires the transmission delay of the service flow through the count value count and the count value sumdelay and inquires the service request f 1 When the transmission delay information of (1) passes through the hash function h 1 To h d For service request f 1 Performing hash for d times to find corresponding d counters, and service request f 1 Query result count of number of data packets s For the minimum value in the count value count of d counter counters, the service request f 1 Query result sumdelay of data packet transmission delay value sum s The minimum of the sumdelay values is calculated for the d counter counters.
Within one period, service request f 1 The average transmission delay between node a and node B is:
Figure BDA0004067835620000081
taking the average transmission delay as the network transmission delay from the node A to the node B;
step S5: emptying bloom filter, counter and hash table
After each measurement cycle is completed, the bloom filter, counter in the sktech algorithm module, and hash table in the stream processor are emptied, ready for the next measurement cycle.
In this embodiment, as shown in fig. 1, the timestamp label is removed after the data packet passes through the node B, so as to avoid unnecessary transmission load.
The invention adds the timestamp label to the data packet, so that when the measurement algorithm is deployed, the node B can obtain the required data packet transmission delay only by providing a small amount of high-speed storage resources, and the abnormal flow is separated according to the time limit threshold value, thereby reducing the resource consumption. Meanwhile, the invention separates the abnormal flow of network performance again through the bloom filter, on one hand, the abnormal flow can be rapidly positioned through the hash table, and on the other hand, after the abnormal value is filtered, the measurement accuracy of the CM Sketch module can be further improved.
While the foregoing describes illustrative embodiments of the present invention to facilitate an understanding of the present invention by those skilled in the art, it should be understood that the present invention is not limited to the scope of the embodiments, but is to be construed as protected by the accompanying claims insofar as various changes are within the spirit and scope of the present invention as defined and defined by the appended claims.

Claims (2)

1. The network transmission delay measuring method of any two nodes in the programmable network is characterized by comprising the following steps:
(1) Packet timestamp label addition and abnormal time delay stream judgment
Service request f initiated to network for a content 1 The network transmission delay from node A to node B needs to be measured, a time stamp label is added to the data packet at node A, and the content of the time stamp label is the entry time stamp t of the data packet reaching the switch A 1 When a data packet is forwarded to the node B, the node B obtains the current timestamp t 2 Then analyzing the time stamp label in the data packet to obtain an entry time stamp t 1 Service request f 1 Transmission delay d between node a and node B l ＝t 2 -t 1 ；
In node B, the transmission delay d is calculated l After that, the transmission delay d l Compared with a time delay threshold delta, the transmission time delay d l Service request f greater than delay threshold delta 1 For transmitting a stream with abnormal performance, namely an abnormal streamEncoding five tuples of the abnormal flow, recording by using a bloom filter, and forwarding the abnormal flow to a flow processor; for transmission delay d l Service request f not greater than delay threshold delta 1 Sending a bloom filter to make further decisions, wherein the bloom filter is deployed in the node B;
(2) The bloom filter further separates the abnormal flow
In bloom filter, for service request f 1 The five-tuple, namely the source IP, the destination IP, the source port, the destination port and the protocol type, are encoded by adopting the same encoding method in the step (1), and compared with the five-tuple encoding of the abnormal flow recorded in the bloom filter, if the five-tuple is the same, the five-tuple is considered as the abnormal flow and is forwarded to a flow processor, if the five-tuple is not the same, the five-tuple is the normal flow and is forwarded to a Sketch algorithm module, wherein the Sketch algorithm module is deployed in a node B;
(3) The stream processor records the abnormal stream and reports the abnormal stream to the controller for positioning
The abnormal flow judged by the time delay threshold delta and the abnormal flow separated by the bloom filter are accurately recorded in the flow processor by using a hash table, the five-tuple and the accumulated time delay of the abnormal flow are reported to the controller according to the measurement period, and the controller can rapidly locate the generated abnormal flow through the hash table;
(4) Network transmission delay measurement by using Sketch algorithm
In the sktch algorithm module, each counter contains two parts: the count value count of 32 bits is used for counting the number of data packets and the count value sumdelay of 32 bits is used for storing the accumulated sum of transmission delay;
when the transmission delay d is calculated l The node B then passes the five-tuple of the data packet through the hash function h 1 To h d D times of hashing are carried out, a counter corresponding to each time of hashing is found to update, and the updating rule is as follows: count=count+1, sumdelay=sumdelay+d l ；
The node B reports the count value count and the count value sumdelay of the counter to the controller according to the measurement period described in the step (3), and the controller transmits the count value count,The count value sumdelay inquires out the transmission delay of the service flow and inquires the service request f 1 When the transmission delay information of (1) passes through the hash function h 1 To h d For service request f 1 Performing hash for d times to find corresponding d counters, and service request f 1 Query result count of number of data packets s For the minimum value in the count value count of d counter counters, the service request f 1 Query result sumdelay of data packet transmission delay value sum s Calculating the minimum value in the sumdelay value for d counter counters;
within one period, service request f 1 The average transmission delay between node a and node B is:
Figure FDA0004067835610000021
taking the average transmission delay as the network transmission delay from the node A to the node B;
(5) Emptying bloom filter, counter and hash table
After each measurement cycle is completed, the bloom filter, counter in the sktech algorithm module, and hash table in the stream processor are emptied, ready for the next measurement cycle.
2. The method of claim 1, wherein the timestamp label is removed after the packet passes through the node B.
CN202310082418.0A 2023-02-08 2023-02-08 Network transmission delay measurement method for any two nodes in programmable network Pending CN116094962A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202310082418.0A CN116094962A (en) 2023-02-08 2023-02-08 Network transmission delay measurement method for any two nodes in programmable network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202310082418.0A CN116094962A (en) 2023-02-08 2023-02-08 Network transmission delay measurement method for any two nodes in programmable network

Publications (1)

Publication Number Publication Date
CN116094962A true CN116094962A (en) 2023-05-09

Family

ID=86211826

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202310082418.0A Pending CN116094962A (en) 2023-02-08 2023-02-08 Network transmission delay measurement method for any two nodes in programmable network

Country Status (1)

Country Link
CN (1) CN116094962A (en)

Similar Documents

Publication Publication Date Title
US11171853B2 (en) Constraint-based event-driven telemetry
US11044204B1 (en) Visibility packets with inflated latency
CN109787833B (en) Network abnormal event sensing method and system
Yu et al. Flowsense: Monitoring network utilization with zero measurement cost
US20190013954A1 (en) Elastic timestamping
US10826813B2 (en) Threshold crossing events for network element instrumentation and telemetric streaming
US11924681B2 (en) Adaptive in-band network telemetry for full network coverage
CN110149239B (en) Network flow monitoring method based on sFlow
US11843541B2 (en) Congestion notification reporting for a responsive network
CN113676376A (en) In-band network telemetering method based on clustering
US9992081B2 (en) Scalable generation of inter-autonomous system traffic relations
Wang et al. A bandwidth-efficient int system for tracking the rules matched by the packets of a flow
CN111726410B (en) Programmable real-time computing and network load sensing method for decentralized computing network
Lei et al. Printqueue: performance diagnosis via queue measurement in the data plane
Ling et al. FullSight: Towards scalable, high-coverage, and fine-grained network telemetry
CN116094962A (en) Network transmission delay measurement method for any two nodes in programmable network
WO2022270766A1 (en) Device and method for automatic packet analysis-based intelligent network management
CN116055362A (en) Two-stage Hash-Sketch network flow measurement method based on time window
CN116436825A (en) Packet loss processing method and system and network equipment
Wu et al. Lossdetection: Real-time packet loss monitoring system for sampled traffic data
CN115037667A (en) Fine-grained network situation awareness and source routing intelligent optimization method and device
CN111800311B (en) Real-time sensing method for decentralized computing state
Andersson Offloading INTCollector Events with P4
Nguyen et al. A Comprehensive P4-based Monitoring Framework for L4S leveraging In-band Network Telemetry
CN110933002B (en) Method and device for realizing switching chip of MPLS in-band detection OAM

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination