CN114143209B - System and method for controlling elastic expansion of channel in network measurement - Google Patents
System and method for controlling elastic expansion of channel in network measurement Download PDFInfo
- Publication number
- CN114143209B CN114143209B CN202111314542.2A CN202111314542A CN114143209B CN 114143209 B CN114143209 B CN 114143209B CN 202111314542 A CN202111314542 A CN 202111314542A CN 114143209 B CN114143209 B CN 114143209B
- Authority
- CN
- China
- Prior art keywords
- control channel
- overload
- load
- migrated
- low
- 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
- H04L41/145—Network analysis or design involving simulating, designing, planning or modelling of a network
-
- 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
- H04L41/12—Discovery or management of network topologies
-
- 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
-
- 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
-
- 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
Abstract
The invention discloses a system and a method for elastically expanding a control channel in network measurement, which can quickly and effectively complete the outward/inward expansion of an overload/low-load control channel by integrating the statistical information of the control channel and network topology, combining an integer linear programming model and establishing constraint conditions. The method is simple, flexible to realize and high in practicability; the influence of overload and low-load control channels on a measurement system can be reduced.
Description
Technical Field
The invention belongs to the field of data acquisition of network measurement in a programmable network, and particularly relates to a system and a method for controlling elastic expansion of a channel in network measurement.
Background
In modern data centers, network measurements have been widely applied and deployed. Network measurements require the control plane to be performed in conjunction with the data plane. In the data plane, measurement tasks are deployed on the switches and process traffic at line rate and extract events (such as packet number) from the traffic. Where each event is associated with a particular stream. Subsequently, the switch of the data plane sends different events to the control plane together through the control channel, and the measurement application of the control plane completes network measurement according to the received events.
However, the existing network measurement scheme has two problems of controlling the channel overload and the channel underload. The highest bandwidth of the control channel is often only 100Gbps, and the throughput of the switch in the data center network can often reach the Tbps level, which causes congestion of the control channel, further causes loss of events, and reduces the accuracy of the measurement system. In addition, on some edge switches, their throughput is often small, and the events generated are relatively few, which results in low load of the control channel and waste of resources.
Disclosure of Invention
The invention aims to provide a system and a method for controlling the elastic expansion of a channel in network measurement, aiming at the defects of the prior art.
The purpose of the invention is realized by the following technical scheme:
a system for elastically expanding a control channel in network measurement, comprising: the system comprises a data acquisition module, an overload/underload control channel detection module, an optimal migration scheme calculation module and a migration strategy implementation module.
(1) And implanting a data acquisition module in an SDN (Software Defined Network) controller of a control plane. The technical scheme for acquiring the statistical information of the control channel by the data acquisition module is as follows:
controlling the real-time load of the channel: the port rate is read directly.
Controlling the transmission delay of the channel: and the controller transmits the LLDP message to the switch and calculates the RTT time of the LLDP message.
Control size of each event stream in a channel: and setting a counter according to the quintuple of the data packets between the event streams, and counting the size of each stream.
Specified transmission delay of event stream: the specified transmission delay of each event stream is set by a user in advance and can be directly obtained; for example, event stream f needs to arrive from the data plane to the control plane within 30 ms.
The technical scheme for acquiring the network topology statistical information by the data acquisition module is as follows: and collecting the load of each link and the transmission delay of each link in the topology. And calculating the time delay and the residual bandwidth of all paths and each path between any two switches by using a depth-first search algorithm in combination with the network topology.
During operation, the data acquisition module sends the control channel statistical information to the overload/low-load control channel detection module, and sends the control channel statistical information and the topology statistical information to the optimal migration scheme calculation module;
(2) Three thresholds theta for overload/low-load control channel detection module top 、Θ bottom And Θ safe The overload and underload of the control channel are detected, the outward migration of the overload control channel is controlled, the extra overhead caused by excessive migration event streams is avoided, and the performance of the measuring system is reduced. For both the overloaded and underloaded control channels, the overloaded/underloaded control channel detection module selects the event stream for which to migrate. Finally, the overload/low-load control channel detection module sends the event flow set to be migrated to the optimal migration scheme calculation module;
(3) And when the optimal migration scheme calculation module receives the control channel statistical information, the topology statistical information and the event stream set to be migrated, establishing an integer linear programming model by taking the control channel statistical information, the topology statistical information and the event stream set to be migrated as input. The output of the model is a migration scheme, determines to which new control channel each event stream to be migrated from the original control channel is, and provides a routing strategy for the event stream to reach the new control channel. The objective function of the model is to minimize the sum of the new propagation delays of the event streams to be migrated.
The constraints of the model are: (i) Any event stream to be migrated cannot be migrated to the original control channel; (ii) all event streams must be completely migrated; (iii) control channels that cannot cause new overloads; (iv) link overload cannot be caused; (v) The transmission delay after all event streams are migrated cannot exceed the specified transmission delay;
(4) And after receiving the migration scheme of the optimal migration scheme calculation module, the migration policy implementation module configures a forwarding table to the switch to realize the migration of the event stream to be migrated.
Further, a data acquisition module is implanted in the SDN controller of the control plane and is used for collecting control channel statistical information and topology statistical information. And calculating the residual bandwidth and the transmission delay of all paths based on the statistical information.
Further, the overload/underload control channel detection module comprises three user pre-fixesDefined threshold theta top 、Θ bottom And Θ safe . When the load of one control channel exceeds theta top The overload/low-load control channel detection module recognizes it as an overload control channel. Similarly, the overload/underload control channel detection module will load below Θ bottom Is identified as a low-load control channel. Theta safe Then the method is used for terminating the outward expansion of the overload control channel; for example, a control channel has a load theta top When it is reduced to theta safe The event stream migrating out of it is stopped in the following. For the overloaded control channel, the overload/underload control channel detection module sequentially selects the event stream with the maximum specified transmission delay to migrate until the load of the event stream is lower than theta safe (ii) a And selecting all event streams for migration in the underloaded control channel.
Further, the constraint conditions of the integer linear programming model established by the optimal migration scheme calculation module are as follows:
(i) Any event stream to be migrated cannot be migrated to the original control channel;
(ii) All event streams must be completely migrated;
(iii) A control channel that cannot cause a new overload;
(iv) Link overload cannot be caused;
(v) The transmission delay after all event stream transitions cannot exceed its specified transmission delay.
A method for flexibly expanding a control channel in network measurement collects statistical information of the control channel, statistical information of an underlying network link and the rate of each measurement task sending flow event during operation. Detecting overload and low-load control channels according to the collected information, performing outward expansion on the overload control channels, namely migrating a part of event streams to reduce the load of the control channels so as to prevent loss of stream events, and performing inward expansion on the low-load control channels, namely migrating all the event streams and closing the control channels so as to improve the utilization rate of network resources. When the outward/inward extension is implemented, it is ensured that no new overload control channel appears, no overload of the link is caused, and the new transmission delay of the migrated event stream does not exceed the specified transmission delay.
The beneficial results of the invention are: the invention integrates the statistical information of the control channel and the network topology, combines an integer linear programming model, and establishes constraint conditions, thereby quickly and effectively finishing the outward/inward expansion of the overload/low-load control channel. The method is simple, flexible to realize and high in practicability; the influence of overload and low-load control channels on the measurement system can be reduced.
Drawings
FIG. 1 is a schematic structural diagram of a control channel elastic expansion system in network measurement;
fig. 2 is a flow chart of the operation of the control channel resilient extension system in network measurement.
Detailed Description
Referring to fig. 1 and fig. 2, the control channel elastic expansion system in network measurement according to the present invention utilizes the control channel and the network topology statistical information collected in real time to process the overloaded/underloaded control channel in time without affecting the network measurement system. In a specific embodiment, the SDN controller selects RYU, and the programmable switch is a switch using a bareboot Tofino chip. The method comprises the following specific steps:
1. the data acquisition module monitors the uplink and downlink rates of each control channel so as to obtain the real-time load of the control channel. For the event streams, the data acquisition module sets a counter according to the quintuple of the corresponding streams, counts the number of the events received by each quintuple in 1s, and further counts the rate of each event stream. And for the time delay of each control channel, when the data acquisition module receives an event, the time delay of the control channel is calculated according to the sending time of the event and the time of receiving the event. And for the time delay of each link, the data acquisition module sends an LLDP message to one end of each link and records the sending time t1, and when the other end receives the LLDP message, the LLDP message is automatically sent to the data acquisition module. The data acquisition module records the event t2 received, so that the time delay of the link can be obtained through (t 2-t1-2 control channel time delay). The data acquisition module can acquire uplink and downlink rates at two ends of a link through the SDN controller so as to acquire the real-time load of the link.
2. Threshold theta defined in advance by three users in overload/low-load control channel detection module top 、Θ bottom And Θ safe Set to 80Gbps, 20Gbps, and 60Gbps, respectively. When selecting the event stream to be migrated, for the overloaded control channel, the overload/underload control channel detection module will always select the event stream with the maximum specified transmission delay for migration until the load is lower than Θ safe (ii) a And selecting all the event streams to be migrated for the underloaded control channel.
3. And the optimal migration scheme calculation module establishes an integer linear programming model. The output of the model is the destination control channel of each event stream to be migrated and its new route. The objective function of the model is to minimize the sum of the new propagation delays of the event streams to be migrated. The constraints of the model are: (i) Any event stream to be migrated cannot be migrated to the original control channel; (ii) all event streams must be completely migrated; (iii) control channels that cannot cause new overloads; (iv) link overload cannot be caused; (v) The transmission delay after all event stream transitions cannot exceed its specified transmission delay. The integer linear programming model is solved by using Gurobi coding.
4. The migration policy enforcement module is implemented by matching action tables in the P4 language when an event stream is migrated. The SDN controller issues the matching action table items to all the switches on the new routing path of the SDN controller, and deployment of the migration strategy is completed.
The invention relates to a method for elastically expanding a control channel in network measurement, which collects statistical information of the control channel, statistical information of an underlying network link and the rate of sending stream events (namely the rate of each event stream) of each measurement task during operation. Detecting overload and low-load control channels according to the acquired information, performing outward expansion on the overload control channels, performing inward expansion on the low-load control channels, namely migrating a part of event streams therein to reduce load, preventing the event streams from being lost due to congestion of the control channels, and performing inward expansion on the low-load control channels, namely migrating all the event streams therein and closing the control channels, so as to improve the utilization rate of network resources. When the outward/inward expansion is implemented, it is ensured that no new overload control channel occurs, no overload of the link occurs, and the new transmission delay of the migrated event stream does not exceed the specified transmission delay.
The above description is only a preferred embodiment of the present invention, and all equivalent changes and modifications made in accordance with the claims of the present invention should be covered by the present invention.
Claims (5)
1. A system for elastically expanding a control channel in network measurement, comprising: the system comprises a data acquisition module, an overload/low-load control channel detection module, an optimal migration scheme calculation module and a migration strategy implementation module;
the data acquisition module is used for collecting control channel statistical information and topology statistical information; sending the control channel statistical information to an overload/low-load control channel detection module, and sending the control channel statistical information and topology statistical information to an optimal migration scheme calculation module; the statistical information of the control channels comprises real-time load, transmission delay and the size of each event stream of each control channel; the network topology information includes: time delay and residual bandwidth of all paths and each path between any two switches;
the overload/low-load control channel detection module selects event streams to be migrated from the overload and low-load control channels, and sends the set of the event streams to be migrated to the optimal migration scheme calculation module;
the optimal migration scheme calculation module is used for establishing an integer linear programming model and taking the received control channel statistical information, topology statistical information and an event flow set to be migrated as the input of the model; the output of the model is a migration scheme which comprises a target control channel of each event flow to be migrated and a new route thereof; the objective function of the model is to minimize the sum of the new transmission delays of the event streams to be migrated;
and the migration strategy implementation module is used for receiving the migration scheme of the optimal migration scheme calculation module, generating corresponding switch configuration and actually deploying the migration scheme in the network.
2. The system of claim 1, wherein a data collection module is embedded in the SDN controller of the control plane, and the data collection module calculates the remaining bandwidth and the transmission delay of all paths based on the statistical information of the control channel and the statistical information of the topology.
3. The system of claim 1, wherein the overload/underload control channel detection module comprises three predetermined thresholds Θ top 、Θ bottom And Θ safe (ii) a When the load of one control channel exceeds theta top The overload/low-load control channel detection module determines the overload/low-load control channel as an overload control channel; the load is lower than theta bottom The control channel of (2) is identified as a low-load control channel; theta safe An outward extension for terminating the overload control channel;
for the overloaded control channel, the overload/underload control channel detection module selects the event stream with the maximum specified transmission delay in the overload control channel to be added into the migrated event stream set until the load of the overload control channel is lower than theta safe ；
For the low-load control channel, the overload/low-load control channel detection module selects all event streams in the low-load control channel to join the migrated event stream set.
4. The system for elastically expanding a control channel in network measurement according to claim 1, wherein the constraint condition of the integer linear programming model established by the optimal migration scheme calculation module is:
(i) Any event stream to be migrated cannot be migrated to the original control channel;
(ii) All event streams must be completely migrated;
(iii) Control channels that cannot cause new overloads;
(iv) Link overload cannot be caused;
(v) The transmission delay after all event stream transitions cannot exceed their specified transmission delay.
5. An elastic expansion method for a control channel in network measurement, which is characterized in that, the elastic expansion system for a control channel in network measurement according to claim 1 collects statistical information of the control channel during operation, the load of each link in topology, the transmission delay of each link and the rate of sending stream events by each measurement task; detecting overload and low-load control channels according to the acquired information, performing outward expansion on the overload control channels, namely transferring a part of event streams to reduce the load of the control channels so as to prevent loss of stream events, and performing inward expansion on the low-load control channels, namely transferring all the event streams and closing the control channels so as to improve the utilization rate of network resources; when the outward/inward expansion is implemented, it is ensured that no new overload control channel occurs, no overload of the link occurs, and the new transmission delay of the migrated event stream does not exceed the specified transmission delay.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111314542.2A CN114143209B (en) | 2021-11-08 | 2021-11-08 | System and method for controlling elastic expansion of channel in network measurement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111314542.2A CN114143209B (en) | 2021-11-08 | 2021-11-08 | System and method for controlling elastic expansion of channel in network measurement |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114143209A CN114143209A (en) | 2022-03-04 |
| CN114143209B true CN114143209B (en) | 2022-10-21 |
Family
ID=80392690
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111314542.2A Active CN114143209B (en) | 2021-11-08 | 2021-11-08 | System and method for controlling elastic expansion of channel in network measurement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114143209B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014000292A1 (en) * | 2012-06-30 | 2014-01-03 | 华为技术有限公司 | Migration method, serving control gateway and system for virtual machine across data centres |
| CN104717137A (en) * | 2013-12-12 | 2015-06-17 | 国际商业机器公司 | Managing data flows in overlay networks |
| WO2017184627A2 (en) * | 2016-04-18 | 2017-10-26 | Nyansa, Inc. | A system and method for network incident identification, congestion detection, analysis, and management |
| CN110505154A (en) * | 2019-07-26 | 2019-11-26 | 北京工业大学 | A kind of interchanger moving method based on improved Dijkstra's algorithm |
| US11016824B1 (en) * | 2017-06-12 | 2021-05-25 | Pure Storage, Inc. | Event identification with out-of-order reporting in a cloud-based environment |
| CN112994937A (en) * | 2021-02-10 | 2021-06-18 | 北京交通大学 | Deployment and migration system of virtual CDN in intelligent fusion identification network |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103685250A (en) * | 2013-12-04 | 2014-03-26 | 蓝盾信息安全技术股份有限公司 | Virtual machine security policy migration system and method based on SDN |
| CN107612771B (en) * | 2017-09-07 | 2020-11-17 | 广东工业大学 | SDN network load balancing method based on dynamic migration |
| US10795715B2 (en) * | 2018-06-28 | 2020-10-06 | At&T Intellectual Property I, L.P. | Cloud oversubscription system |
| CN110765595B (en) * | 2019-10-10 | 2022-12-09 | 内蒙古农业大学 | SDN data center network flow scheduling method based on multilayer virtual topology energy saving |
-
2021
- 2021-11-08 CN CN202111314542.2A patent/CN114143209B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2014000292A1 (en) * | 2012-06-30 | 2014-01-03 | 华为技术有限公司 | Migration method, serving control gateway and system for virtual machine across data centres |
| CN104717137A (en) * | 2013-12-12 | 2015-06-17 | 国际商业机器公司 | Managing data flows in overlay networks |
| WO2017184627A2 (en) * | 2016-04-18 | 2017-10-26 | Nyansa, Inc. | A system and method for network incident identification, congestion detection, analysis, and management |
| US11016824B1 (en) * | 2017-06-12 | 2021-05-25 | Pure Storage, Inc. | Event identification with out-of-order reporting in a cloud-based environment |
| CN110505154A (en) * | 2019-07-26 | 2019-11-26 | 北京工业大学 | A kind of interchanger moving method based on improved Dijkstra's algorithm |
| CN112994937A (en) * | 2021-02-10 | 2021-06-18 | 北京交通大学 | Deployment and migration system of virtual CDN in intelligent fusion identification network |
Non-Patent Citations (2)
| Title |
|---|
| Bassem Sellami ; Akram Hakiri ; Sadok Ben Yahia ; Pascal Berthou.Deep Reinforcement Learning for Energy-Efficient Task Scheduling in SDN-based IoT Network.《2020 IEEE 19th International Symposium on Network Computing and Applications (NCA)》.2021, * |
| 大规模服务覆盖网拓扑设计;张栋，吴春明，姜明，熊伟;《电子与信息学报》;20100415;第35卷(第3期) * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114143209A (en) | 2022-03-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108833279B (en) | Method for multi-constraint QoS routing based on service classification in software defined network | |
| US20210176177A1 (en) | Network control apparatus and network control method | |
| CN105357075B (en) | Flow monitoring system and method based on software defined network | |
| CN108512760A (en) | The method for routing of QoS of survice is ensured based on SDN | |
| CN109547341B (en) | Load sharing method and system for link aggregation | |
| CN108718283B (en) | TCP congestion control method for centralized end network coordination in data center network | |
| US20070115846A1 (en) | Method for controlling data throughput in a storage area network | |
| EP4024778A1 (en) | Method for determining required bandwidth for data stream transmission, and devices and system | |
| US20200136973A1 (en) | Dynamically balancing traffic in a fabric using telemetry data | |
| KR20150036418A (en) | Delay-based traffic rate control in networks with central controllers | |
| CN111970759B (en) | Delay adjustment method and device for end-to-end service, storage medium and electronic device | |
| JP6092409B2 (en) | Method and apparatus for evaluating wireless network capillary performance | |
| EP2087661A1 (en) | A method for improved congestion detection and control in a wireless telecommunications systems. | |
| JP2006506845A (en) | How to select a logical link for a packet in a router | |
| Zhou et al. | Analysis and implementation of packet preemption for time sensitive networks | |
| CN110557333A (en) | method and system for controlling and guaranteeing quality of service of software defined network | |
| CN111935022A (en) | Flow table consistency updating method in software defined network | |
| CN113746700B (en) | Elephant flow rapid detection method and system based on probability sampling | |
| CN114143209B (en) | System and method for controlling elastic expansion of channel in network measurement | |
| US9553828B1 (en) | System and method for collision detection and avoidance for network communications | |
| JP2023509426A (en) | System and method for real-time monitoring and optimization of mobile networks | |
| US20120311169A1 (en) | Network relay apparatus | |
| CN104902570B (en) | A kind of dynamics of channels configuration method and device | |
| Çakmak et al. | A Review: Active queue management algorithms in mobile communication | |
| WO2021078231A1 (en) | Location awareness-based network intermediate device |
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 |