Summary
Asia-Pacific Network Operations and Management Symposium
2022
Session Number:PS4
Session:
Number:PS4-11
GENEVE@TEIN: A Sophisticated Tunneling Technique for Communication Between OpenStack-Based Multiple Clouds at TEIN.
Mir Muhammad Suleman Sarwar, Javier Jose Diaz Rivera, Afaq Muhammad, Wang-Cheol Song,
pp.-
Publication Date:2022/09/28
Online ISSN:2188-5079
PDF download
Summary:
Multiple clouds need to share resources as a service for various reasons, such as overcoming single points of failure or reducing latency. Tunneling serves as the mechanism for multiple clouds to share their resources. Generic Network Virtualization Encapsulation (GENEVE) tunnel has the advantage of a flexible header over other L2 tunnels such as GRE or VXLAN. This enables the GENEVE tunnel to transfer a larger payload in case of smaller header size. The service must be acquired from an optimal Cloud and Instance. An optimal Cloud has the best network performance forecasted. An optimal Instance is forecasted not to be overutilized during resource sharing. For this purpose, monitoring tools must be deployed for metering Instances and network links. Data obtained from monitoring tools can be utilized by Machine Learning for Instances and network performance forecasting. Initially, two clouds are deployed at Trans-Eurasia Internetworking (TEIN), one in Malaysia and the other in South Korea. Monitoring tools are deployed for network and Instance performance monitoring. GENEVE tunnel is then deployed on OVS bridges, after which Instances of a cloud can access Instances of the other Cloud for resource sharing.