Detta är ett uppsatsförslag hämtat från Nationella Exjobb-poolen. Klicka här för att komma tillbaka till samtliga exjobbsförslag.
Measurement of node availability in a distributed network
Degoo is developing a P2P Backup system which has its roots in the Master’s Thesis “CollabBackup: A Peer-to-Peer backup system focused on storage efficiency”. There’s still some work needed before the system is ready for the commercial market. This Master’s Thesis is an important part of that.
The goal of the thesis is to investigate and implement a mathematical model which calculates the expected availability of a given computer node in a network. One of the reasons this task is not trivial is that all measurements are done by other nodes in the network and you cannot assume that any of these nodes are trustworthy. You therefore have to calculate an aggregate value based on the results from all the reporting nodes. One possible solution to this is to use a Markov-model to calculate the availability of the nodes. This would be quite similar to Google’s original PageRank algorithm, where the trustworthiness of a site was based on how many other sites were linking to it and how trustworthy they were. In other words: you would put a bias on the reported data based on the reporting node’s own availability. This assumes that trustworthiness is equivalent to availability.
Another challenge is that a node’s availability might be time dependent (e.g. a node might be turned off at night) and you therefore will have to take the time/date of the measurements in consideration when you make future predictions.
The goal of this thesis is therefore to:
* Investigate potential solutions which would enable accurate prediction of all nodes’ respective availability at a large scale, while still be resistant to malicious nodes reporting incorrect data.
* Implement the model in Degoo’s P2P Backup system.
* Write the final report.
Informationen om uppsatsförslag är hämtad från Nationella Exjobb-poolen.