Graph Modeling for OpenFlow Switch Monitoring
Network monitoring allows network administrators to facilitate network activities and to resolve issues in a timely fashion. Monitoring techniques in software-defined networks are either (i) active, where probing packets are sent periodically, or (ii) passive, where traffic statistics are collected...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2023-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/10213403/ |
| _version_ | 1797243535512043520 |
|---|---|
| author | Ali Malik Ruairi De Frein |
| author_facet | Ali Malik Ruairi De Frein |
| author_sort | Ali Malik |
| collection | DOAJ |
| description | Network monitoring allows network administrators to facilitate network activities and to resolve issues in a timely fashion. Monitoring techniques in software-defined networks are either (i) active, where probing packets are sent periodically, or (ii) passive, where traffic statistics are collected from the network forwarding elements. The centralized nature of software-defined networking implies the implementation of monitoring techniques imposes additional overhead on the network controller. We propose Graph Modeling for OpenFlow Switch Monitoring (GMSM), which is a lightweight monitoring technique. GMSM constructs a flow-graph overview using two types of asynchronous OpenFlow messages: <monospace>packet-in</monospace> and <monospace>flow-removed</monospace>, which improve monitoring and decision making. It classifies new flows based on the class of service. Experimental findings suggest that using GMSM leads to a decrease in network overhead resulting from the communication between the controller and the switches, with a reduction of 5.7% and 6.7% compared to state-of-the-art approaches. GMSM reduces the controller’s CPU utilization by more than 2% compared to other monitoring methods. Overhead reduction comes with a slight reduction of approximately 0.17 units in the estimation accuracy of links utilization because GMSM allows the user to monitor the network subject to a selected class of service, as opposed to having an exact view of the network utilization. |
| first_indexed | 2024-04-24T18:56:40Z |
| format | Article |
| id | doaj.art-cc59ae97a26f4322bc28433f43e5a780 |
| institution | Directory Open Access Journal |
| issn | 2169-3536 |
| language | English |
| last_indexed | 2024-04-24T18:56:40Z |
| publishDate | 2023-01-01 |
| publisher | IEEE |
| record_format | Article |
| series | IEEE Access |
| spelling | doaj.art-cc59ae97a26f4322bc28433f43e5a7802024-03-26T17:34:55ZengIEEEIEEE Access2169-35362023-01-0111845438455310.1109/ACCESS.2023.330384710213403Graph Modeling for OpenFlow Switch MonitoringAli Malik0https://orcid.org/0000-0002-2866-0743Ruairi De Frein1https://orcid.org/0000-0002-3912-1470School of Electrical and Electronic Engineering, Technological University Dublin, Dublin 7, IrelandSchool of Electrical and Electronic Engineering, Technological University Dublin, Dublin 7, IrelandNetwork monitoring allows network administrators to facilitate network activities and to resolve issues in a timely fashion. Monitoring techniques in software-defined networks are either (i) active, where probing packets are sent periodically, or (ii) passive, where traffic statistics are collected from the network forwarding elements. The centralized nature of software-defined networking implies the implementation of monitoring techniques imposes additional overhead on the network controller. We propose Graph Modeling for OpenFlow Switch Monitoring (GMSM), which is a lightweight monitoring technique. GMSM constructs a flow-graph overview using two types of asynchronous OpenFlow messages: <monospace>packet-in</monospace> and <monospace>flow-removed</monospace>, which improve monitoring and decision making. It classifies new flows based on the class of service. Experimental findings suggest that using GMSM leads to a decrease in network overhead resulting from the communication between the controller and the switches, with a reduction of 5.7% and 6.7% compared to state-of-the-art approaches. GMSM reduces the controller’s CPU utilization by more than 2% compared to other monitoring methods. Overhead reduction comes with a slight reduction of approximately 0.17 units in the estimation accuracy of links utilization because GMSM allows the user to monitor the network subject to a selected class of service, as opposed to having an exact view of the network utilization.https://ieeexplore.ieee.org/document/10213403/Software-defined networkingOpenFlowmonitoringoverheadutilization |
| spellingShingle | Ali Malik Ruairi De Frein Graph Modeling for OpenFlow Switch Monitoring IEEE Access Software-defined networking OpenFlow monitoring overhead utilization |
| title | Graph Modeling for OpenFlow Switch Monitoring |
| title_full | Graph Modeling for OpenFlow Switch Monitoring |
| title_fullStr | Graph Modeling for OpenFlow Switch Monitoring |
| title_full_unstemmed | Graph Modeling for OpenFlow Switch Monitoring |
| title_short | Graph Modeling for OpenFlow Switch Monitoring |
| title_sort | graph modeling for openflow switch monitoring |
| topic | Software-defined networking OpenFlow monitoring overhead utilization |
| url | https://ieeexplore.ieee.org/document/10213403/ |
| work_keys_str_mv | AT alimalik graphmodelingforopenflowswitchmonitoring AT ruairidefrein graphmodelingforopenflowswitchmonitoring |