Hide metadata

dc.contributor.authorHapnes, Erlend Kristoffer Hatløy
dc.date.accessioned2023-08-24T22:03:28Z
dc.date.available2023-08-24T22:03:28Z
dc.date.issued2023
dc.identifier.citationHapnes, Erlend Kristoffer Hatløy. A Large-Scale Measurement Study of the IPv6 Flow Label. Master thesis, University of Oslo, 2023
dc.identifier.urihttp://hdl.handle.net/10852/103926
dc.description.abstractCongestion control coupling methods rely on knowing if multiple flows between the same host-pair share the same bottleneck. However, IPv4 load balancing hash algorithms traditionally use data from the transport layer header for flow identification, making such a guarantee impossible. We propose a novel method of controlling packet path behaviour using only IPv6 network-layer header data, and perform measurement studies from vantage points stationed around the globe to verify our proposed methodology. We identify challenges with traceroute measurement studies in the face of Equal-Cost Multi-Path (ECMP) and Link Aggregation Group (LAG) load balancing. Finally, we devise and execute an extensive measurement study to investigate whether the IPv6 flow label is completely carried to its destination and whether it can influence the behaviour of load balancing algorithms using a Paris traceroute-based implementation.eng
dc.language.isoeng
dc.subjecttcp
dc.subjectipv6
dc.subjectcongestion control
dc.subjecttraceroute
dc.titleA Large-Scale Measurement Study of the IPv6 Flow Labeleng
dc.typeMaster thesis
dc.date.updated2023-08-25T22:04:05Z
dc.creator.authorHapnes, Erlend Kristoffer Hatløy
dc.type.documentMasteroppgave


Files in this item

Appears in the following Collection

Hide metadata