Hide metadata

dc.contributor.authorGembala, Bartosz
dc.date.accessioned2018-08-28T22:01:32Z
dc.date.available2018-08-28T22:01:32Z
dc.date.issued2018
dc.identifier.citationGembala, Bartosz. Configuration of Network Parameters in Operating Systems using Machine Learning. Master thesis, University of Oslo, 2018
dc.identifier.urihttp://hdl.handle.net/10852/63899
dc.description.abstractBy default, the Linux network stack is not configured for high-speed large file transfer. The reason behind this is to merely save memory resources. We may tune the Linux network stack by increasing network buffers size for high-speed networks that connect server systems in order to handle more network packets. There are also several other TCP/IP parameters that can be tuned in an OS. In this thesis, the goal is to make a system which learns from the history of the network traffic and leverages this knowledge to optimize the current performance by adjusting the parameters. This can be done for a standard Linux kernel using sysctl or /proc. For a VM, virtually any type of OS can be installed and an image can swiftly be compiled and deployed. By being a sandboxed environment, risky configurations can be performed without the danger of harming the system. Different scenarios for network parameter configurations are thoroughly tested, and an increase of at least 65% throughput speed is achieved compared to the default configuration.eng
dc.language.isoeng
dc.subjectgenetic algorithm
dc.subjectTCP
dc.subjectUDP
dc.subjectconfiguration optimization
dc.subjectVM
dc.subjectNIC
dc.subjectthroughput
dc.subjectmachine learning
dc.subjectparameters
dc.subjectnetwork
dc.titleConfiguration of Network Parameters in Operating Systems using Machine Learningeng
dc.typeMaster thesis
dc.date.updated2018-08-28T22:01:32Z
dc.creator.authorGembala, Bartosz
dc.identifier.urnURN:NBN:no-66435
dc.type.documentMasteroppgave
dc.identifier.fulltextFulltext https://www.duo.uio.no/bitstream/handle/10852/63899/1/Bartosz_Gembala.pdf


Files in this item

Appears in the following Collection

Hide metadata