CN111371612B - Network equipment port availability evaluation method - Google Patents
Network equipment port availability evaluation method Download PDFInfo
- Publication number
- CN111371612B CN111371612B CN202010134919.5A CN202010134919A CN111371612B CN 111371612 B CN111371612 B CN 111371612B CN 202010134919 A CN202010134919 A CN 202010134919A CN 111371612 B CN111371612 B CN 111371612B
- Authority
- CN
- China
- Prior art keywords
- port
- utilization rate
- queue length
- availability
- predicted value
- 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.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L41/00—Arrangements for maintenance, administration or management of data switching networks, e.g. of packet switching networks
- H04L41/14—Network analysis or design
-
- 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
- H04L43/0805—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters by checking availability
-
- 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
Abstract
The invention discloses a method for evaluating the availability of a network equipment port, which relates to the field of communication and comprises a port availability evaluating device, wherein the evaluating device comprises a collector, a processor, a transmitter and a receiver, all port availability data on the equipment can be efficiently collected in real time by arrangement on each line card of the network equipment, the port availability data comprises the length of a port cache queue (the number of messages to be sent cached by a port) and the port flow utilization rate (the percentage of the data rate of the port passing through the port to the capacity of the port), and in consideration of the transmission time of the availability data on the equipment, the method uses an Exponential Weighted Moving Average (EWMA) method to predict the size of a data value of the port after the transmission of the availability data is finished, and then selects a better port to forward the data according to the predicted value.
Description
Technical Field
The invention relates to the field of communication, in particular to a network equipment port availability evaluation method.
Background
The port availability of a network device is affected by the size of the data traffic currently being assumed by the port, in addition to the physical components. For example, there are three ports a, B, and C with a rate of 1Gbits/s, where the ports B and C simultaneously transmit data packets to the port a at a line speed, and at this time, the packet rate arriving at the port a is 2Gbits/s, and a packet with 1Gbits/s is discarded at the port a, and the availability of the port a is poor, and if there is another port D that needs to forward a packet to the port a, the packet will be discarded at the same port a (assuming that the priority of the packet transmitted by B, C, D is the same); for forwarding with a single ingress and a single egress, the dropping cannot be avoided, but for forwarding with multiple egress options, such as aggregation of one link for egress, the dropping can be avoided.
Link aggregation techniques may bind multiple physical ports into one logical port, i.e., a link aggregation port, commonly referred to as trunk, which may increase the forwarded data bandwidth while increasing reliability. For the packet forwarded to the link aggregation port, which physical port in the aggregation port is specifically selected, there are two commonly used methods at present:
1. and performing hash operation by using the characteristics of the message, including mac, IP, port number and the like, and using an operation result as a physical port selection basis. This method is simple and efficient, but does not consider at all how the selected physical port is available, whether it is congested or not.
2. And selecting the physical port by using hash operation, judging the availability of the physical port before selecting the physical port, and selecting one of other physical ports in the aggregation port again if the availability of the selected physical port is poor according to port information which is periodically acquired by a control surface of the equipment. Compared with the method 1, the method is more reasonable, the control plane cpu is very simple to implement, but the efficiency of collecting the port information by relying on the control plane is low, and the availability of the port cannot be reflected in real time.
Disclosure of Invention
The invention aims to provide a network equipment port availability evaluation method, which can efficiently collect availability data of all ports on the equipment in real time, wherein the availability data comprises port cache queue length (the number of messages to be sent cached by the ports) and port flow utilization rate (the percentage of the data rate of the current passing of the ports to the capacity of the ports), and in consideration of the time required for the transmission of the availability data on the equipment, the invention uses an exponential weighted moving average method (EWMA) to predict the size of a data value of the port after the transmission of the availability data is finished, and then selects a better port for data forwarding according to the predicted value.
A method for evaluating the availability of network equipment ports comprises a port availability evaluating device, wherein the port availability evaluating device comprises a collector, a processor, a transmitter and a receiver, and is deployed on each line card of network equipment;
the system comprises a collector, a processor, a transmitter and a receiver, wherein the collector is used for collecting port flow utilization rate and queue length data on a line card and recording historical flow utilization rate and queue length data of the port, the processor is used for calculating the flow utilization rate and the queue length predicted value of the port at the next moment by using an exponential weighted moving average method according to the historical port flow utilization rate and the queue length data of the port at the line card, the transmitter is used for transmitting the calculated flow utilization rate and the queue length predicted value of the port at the next moment to all the line cards, and the receiver is used for receiving the flow utilization rate and the queue length predicted value of the port at the next moment by forwarding planes of all the line cards and storing the flow utilization rate and the queue length predicted value into the storage of all the line cards;
the port availability evaluation method comprises the following steps:
the method comprises the following steps: the forwarding plane of the line card regularly collects and stores the flow utilization rate and the port queue length of all ports of the line card;
step two: the line card forwarding plane calculates a predicted value of the port flow utilization rate and the queue length at the next moment by using an exponential weighted moving average method according to the stored historical port flow utilization rate and the queue length and the currently acquired real-time port flow utilization rate and the queue length, and sends the predicted value to all line cards;
step three: and the forwarding planes of all the line cards receive the port flow utilization rate and the queue length predicted value sent by the forwarding plane of the line card, and supply the link aggregation load for sharing or equipment monitoring.
Preferably, the port availability evaluation method is not only applicable to a distributed architecture but also applicable to a centralized architecture, and the equipment in the centralized architecture is equivalent to a cable card board of the distributed architecture equipment.
Preferably, in the first step, the current real-time traffic utilization rate of the port and the queue length acquisition period default to 100us, and the period can be provided for a user to configure.
Preferably, in the second step, the algorithm for calculating the queue length and the rate prediction value of the port at the next time is as follows:
wherein
Is a predicted value at time t, Yt-1Is the observed value at the time t-1,
is a predicted value at time t-1, α (0)<α<1) For the smoothing factor, it determines the decay rate 1- α of the historical data weight, and by default, α is 0.4, which can be set by the user according to the specific network.
Preferably, in the second step, the line card forwarding plane sends the port traffic utilization and the predicted value of the queue length to all line cards, including the line card itself.
Preferably, in the third step, the threshold for determining the port availability by link aggregation may be set by the user, and if the threshold is exceeded, it is determined that the availability is poor, the device reselects the port.
The invention has the advantages that: the method can efficiently collect all port availability data on the equipment in real time, wherein the port availability data comprises port cache queue length (the number of messages to be sent cached by a port) and port flow utilization rate (the percentage of the data rate of the port passing through the port to the capacity of the port), and in consideration of the time required for the availability data to be transmitted on the equipment, the method uses an Exponential Weighted Moving Average (EWMA) method to predict the size of a data value of the port after the availability data is transmitted, and then selects a better port to forward the data according to the predicted value.
Drawings
FIG. 1 is a schematic diagram of the present invention applied in a link aggregation scenario;
FIG. 2 is a diagram of a network device architecture to which the present invention relates;
FIG. 3 is a flow chart of a processing device on a line card of a device according to the present invention;
FIG. 4 is a flow chart of the evaluation method of the present invention.
Detailed Description
In order to make the technical means, the creation characteristics, the achievement purposes and the effects of the invention easy to understand, the invention is further described with the specific embodiments.
As shown in fig. 1 to 4, a method for evaluating port availability of a network device includes a port availability evaluating apparatus, which includes a collector, a processor, a transmitter, and a receiver, and is deployed on each line card of the network device;
the system comprises a collector, a processor, a transmitter and a receiver, wherein the collector is used for collecting port flow utilization rate and queue length data on a line card and recording historical flow utilization rate and queue length data of the port, the processor is used for calculating the flow utilization rate and the queue length predicted value of the port at the next moment by using an exponential weighted moving average method according to the historical port flow utilization rate and the queue length data of the port at the line card, the transmitter is used for transmitting the calculated flow utilization rate and the queue length predicted value of the port at the next moment to all the line cards, and the receiver is used for receiving the flow utilization rate and the queue length predicted value of the port at the next moment by forwarding planes of all the line cards and storing the flow utilization rate and the queue length predicted value into the storage of all the line cards;
the port availability evaluation method comprises the following steps:
the method comprises the following steps: the forwarding plane of the line card regularly collects and stores the flow utilization rate and the port queue length of all ports of the line card;
step two: the line card forwarding plane calculates a predicted value of the port flow utilization rate and the queue length at the next moment by using an exponential weighted moving average method according to the stored historical port flow utilization rate and the queue length and the currently acquired real-time port flow utilization rate and the queue length, and sends the predicted value to all line cards;
step three: and the forwarding planes of all the line cards receive the port flow utilization rate and the queue length predicted value sent by the forwarding plane of the line card, and supply the link aggregation load for sharing or equipment monitoring. The port availability evaluation method and device on the network equipment can efficiently collect availability data of all ports on the equipment in real time, and the data can be collected on all cable card boards of the network equipment, no matter whether the equipment is centralized box-type equipment or distributed frame-type equipment or cluster system equipment. The relevant devices used by the method are deployed in a forwarding plane of the equipment, are more real-time and efficient compared with the traditional method for collecting data through a control plane, and perform data prediction through a prediction algorithm by utilizing the short-term similarity of the flow, so that the method is closer to an actual network flow model.
The port availability evaluation method is not only suitable for a distributed architecture but also suitable for a centralized architecture, and equipment of the centralized architecture is equivalent to a cable card board of equipment of the distributed architecture.
In the first step, the current real-time traffic utilization rate of the port and the queue length acquisition period are defaulted to 100us, and the period can be provided for a user to configure. Approximately equivalent to the time for a 1Gbits/s port to transmit 10 1500 byte data frames. The recorded data are: 1. historical traffic utilization and queue length 2 of port, current real-time traffic utilization and queue length of port
In the second step, the algorithm for calculating the queue length and the rate prediction value of the port at the next moment is as follows:
wherein
Is a predicted value at time t, Yt-1Is the observed value at the time t-1,
is a predicted value at time t-1, alpha (0)<α<1) For the smoothing factor, it determines the decay rate 1- α of the historical data weight, and by default, α is 0.4, which can be set by the user according to the specific network.
In the second step, the line card forwarding plane sends the port traffic utilization rate and the queue length prediction value to all the line cards, including the line card itself.
In the third step, the threshold value for judging the availability of the port by link aggregation can be set by the user, and if the threshold value is exceeded, the availability is judged to be poor, the device reselects the port. Such as setting port traffic utilization above 90% and port queue length above 10 as an indication of poor availability.
The specific implementation mode and principle are as follows:
the network device port availability evaluation method is used in a link aggregation scenario, as shown in fig. 1, a distributed device has a plurality of line card boards, and three line card boards have ports a, B, and C, respectively. Binding in a link aggregation, assuming that data is coming in from any line card, such as line card 1, and needs to be sent out from the link aggregation, the device needs to select a port from a, B, and C for the data to send, if the selected port is a at this time; because the collector automatically acquires the historical flow utilization rate and the queue length of the port and the current real-time flow utilization rate and the queue length of the port every 100us, the processor calculates the flow utilization rate and the queue length predicted value of the port at the next moment by using an exponential weighted moving average method according to the historical flow utilization rate and the queue length data of the port, and the method specifically adopts the following formula:
calculating the queue length M and the flow utilization rate N of the port A at the next moment;
the sender sends the calculated flow utilization rate N and the queue length predicted value M of the port at the next moment to all the line cards, and the line cards receive the flow utilization rate N and the queue length predicted value of the port at the next moment through the receivers and store the flow utilization rate N and the queue length predicted value into the storage of all the line cards;
if the port traffic utilization rate exceeds 90% and the port queue length exceeds 10, which is an index of poor availability, is set in advance, if M >10 or N > 90%, the port A is determined to be poor in availability, and then the port B is selected again;
and after the port B is selected, judging the port B by using the same evaluation method, if the judgment result of the port B is poor in availability, judging the port C by using the same evaluation method, and if the evaluation result of the port C is that the port flow utilization rate is lower than 90% or the port queue length is smaller than 10, judging that the result of the port C is good in availability, and selecting the port C for forwarding.
Based on the above, the invention can efficiently collect all port availability data on the device in real time, including the port cache queue length (the number of messages to be sent cached by the port) and the port traffic utilization rate (the percentage of the data rate currently passed by the port to the port capacity), and in consideration of the time required for the availability data to be transmitted on the device, the invention uses the exponential weighted moving average method (EWMA) to predict the data value size of the port after the availability data transmission is finished, and then selects a better port to forward the data according to the predicted value.
It will be appreciated by those skilled in the art that the invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. The embodiments disclosed above are therefore to be considered in all respects as illustrative and not restrictive. All changes which come within the scope of or equivalence to the invention are intended to be embraced therein.
Claims (6)
1. A method for evaluating the availability of a network equipment port is applied to a port availability evaluating device, wherein the evaluating device comprises a collector, a processor, a transmitter and a receiver, and is deployed on each line card of the network equipment;
the system comprises a collector, a processor, a transmitter and a receiver, wherein the collector is used for collecting port flow utilization rate and queue length on a line card and recording historical flow utilization rate and queue length of the port, the processor is used for calculating flow utilization rate and queue length predicted value of the port at the next moment by using an exponential weighted moving average method according to the historical flow utilization rate and queue length of the port on the line card, the transmitter is used for transmitting the calculated flow utilization rate and queue length predicted value of the port at the next moment to all the line cards, and the receiver is used for receiving the flow utilization rate and queue length predicted value of the port at the next moment by forwarding planes of all the line cards and storing the flow utilization rate and queue length predicted values in storage of all the line cards;
the port availability evaluation method comprises the following steps:
the method comprises the following steps: the forwarding plane of the line card regularly collects and stores the flow utilization rate and queue length of all ports of the line card;
step two: the line card forwarding plane calculates to obtain a port flow utilization rate and a queue length predicted value at the next moment by using an exponential weighted moving average method according to the stored historical flow utilization rate and the queue length of the port and the currently acquired real-time flow utilization rate and the queue length of the port, and sends the port flow utilization rate and the queue length predicted value at the next moment to all line cards;
step three: and the forwarding planes of all the line cards receive the port flow utilization rate and the queue length predicted value at the next moment sent by the forwarding plane of the line card, and supply the link aggregation load for sharing use or equipment monitoring use.
2. The method of claim 1, wherein the method comprises: the port availability evaluation method is not only suitable for a distributed architecture but also suitable for a centralized architecture, and equipment of the centralized architecture is equivalent to a cable card board of equipment of the distributed architecture.
3. The method of claim 1, wherein the method comprises: in the first step, the current real-time traffic utilization rate of the port and the queue length acquisition period are defaulted to 100us, and the period can be provided for a user to configure.
4. The method of claim 1, wherein the method comprises: in the second step, the algorithm for calculating the port traffic utilization rate and the queue length predicted value at the next moment is as follows:
wherein
Is a predicted value at the time t,
is the observed value at the time t-1,
is a predicted value at the time t-1,
（0<
<1) for the smoothing coefficient, it determines the decay rate 1 of the historical data weight
By default, get
At 0.4, the user can set up according to the specific network.
5. The method of claim 1, wherein the method comprises: in the second step, the line card forwarding plane sends the port traffic utilization rate and the queue length prediction value to all the line cards, including the line card itself.
6. The method of claim 1, wherein the method comprises: in the third step, the threshold value for judging the availability of the port by link aggregation can be set by the user, and if the threshold value is exceeded, the availability is judged to be poor, the device reselects the port.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010134919.5A CN111371612B (en) | 2020-03-02 | 2020-03-02 | Network equipment port availability evaluation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010134919.5A CN111371612B (en) | 2020-03-02 | 2020-03-02 | Network equipment port availability evaluation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111371612A CN111371612A (en) | 2020-07-03 |
| CN111371612B true CN111371612B (en) | 2022-06-14 |
Family
ID=71210256
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010134919.5A Active CN111371612B (en) | 2020-03-02 | 2020-03-02 | Network equipment port availability evaluation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111371612B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102045253A (en) * | 2010-12-06 | 2011-05-04 | 杭州华三通信技术有限公司 | Method and device for sending video surveillance service messages in link aggregation network |
| CN102223300A (en) * | 2011-06-09 | 2011-10-19 | 武汉烽火网络有限责任公司 | Transmission control method for multimedia data in network equipment |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8995277B2 (en) * | 2012-10-30 | 2015-03-31 | Telefonaktiebolaget L M Ericsson (Publ) | Method for dynamic load balancing of network flows on LAG interfaces |
-
2020
- 2020-03-02 CN CN202010134919.5A patent/CN111371612B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102045253A (en) * | 2010-12-06 | 2011-05-04 | 杭州华三通信技术有限公司 | Method and device for sending video surveillance service messages in link aggregation network |
| CN102223300A (en) * | 2011-06-09 | 2011-10-19 | 武汉烽火网络有限责任公司 | Transmission control method for multimedia data in network equipment |
Non-Patent Citations (1)
| Title |
|---|
| 一种基于自相似流量速率估计的主动队列管理算法;汪岩等;《计算机应用》;20051230;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111371612A (en) | 2020-07-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11451481B2 (en) | Network control apparatus and network control method | |
| US10333848B2 (en) | Technologies for adaptive routing using throughput estimation | |
| US8693489B2 (en) | Hierarchical profiled scheduling and shaping | |
| CN109756421B (en) | Congestion control system and method based on OpenFlow technology | |
| CN111355669B (en) | Method, device and system for controlling network congestion | |
| US20070153683A1 (en) | Traffic rate control in a network | |
| EP3479530B1 (en) | Message schema control | |
| CN112887217B (en) | Control data packet sending method, model training method, device and system | |
| CN111314243A (en) | LoRa network QoS scheduling management method supporting complex service data transmission | |
| JP5374711B2 (en) | Network system, connection device, and data transmission method | |
| JP2013066062A (en) | Remote monitoring system, network interconnection device, and communication control method | |
| CN115473855B (en) | Network system and data transmission method | |
| CN115794407A (en) | Computing resource allocation method and device, electronic equipment and nonvolatile storage medium | |
| WO2018004978A1 (en) | Technologies for adaptive routing using network traffic characterization | |
| CN111371612B (en) | Network equipment port availability evaluation method | |
| CN116827867A (en) | Low-delay congestion flow identification method based on data center network | |
| Sonmez et al. | SUIT: A cross layer image transport protocol with fuzzy logic based congestion control for wireless multimedia sensor networks | |
| CN113747254B (en) | Video stream scheduling method and system based on in-band network telemetry | |
| CN106911531A (en) | A kind of data packetloss rate statistical method of many distribution scenes | |
| CN109257302B (en) | Packet scattering method based on packet queuing time | |
| JP5128556B2 (en) | Traffic information collecting apparatus, traffic information collecting method and program thereof | |
| Pakulova et al. | Simulation of priority multi-path data transmission in the AnyLogic environment | |
| CN115733755A (en) | Data center transmission control system and method capable of filling network bandwidth | |
| CN102082735B (en) | Method for managing passive queue by abandoning head for N times | |
| Sun | An active congestion detection mechanism based on network tomography |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |