You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am working here on my bachelor thesis under supervision of the Net. Sec. Group from ETH Zürich. The goal of the thesis is to analyse the currently implemented beaconing algorithm in terms of bandwidth usage, i.e. analysing its bandwidth overhead and comparing this to possible new strategies.
My intention with this issue is to propose adding bandwidth metrics to the control-service for beaconing and path operations, and announce that I'm working on this actively to avoid duplication of work.
SCION is already using go-grpc-middleware V1 [1] to monitor certain metrics. go-grpc-prometheus [2] is the specific library enabling metric monitoring for gRPC. This library is missing the needed functionality for bandwidth monitoring, but is a good basis to start with. There was already an issue [3] and a PR [4] for integrating bandwidth monitoring to go-grpc-prometheus. The work was put on hold by their maintainers to move now everything to go-grpc-middleware V2.
The idea is to use a different fork from go-grpc-prometheus and integrate the needed functionality to SCION similar to what was already be done in the PR.
Great to see that you are working in this direction.
Currently, there are no efforts underway from our sides with regards to bandwidth monitoring of the control plane and it's also not a priority right now.
Given that you want to use an experimental feature (not even merged yet) I suggest you integrate this in your own fork of the repository. We certainly don't feel comfortable integrating this in the main repository at this point in time.
I'm collaborating with @JST1703 on this. That's definitely the plan. Primary reason for this issue is to avoid duplicate work, but this seems cleared up now.
@JST1703 hope your thesis turned out fine. Is this issue still needed? If not please close it.
oncilla
changed the title
Bandwidth Metrics - measuring badnwidth overhead in beaconing process for bachelor thesis
Bandwidth Metrics - measuring bandwidth overhead in beaconing process for bachelor thesis
Oct 26, 2021
I am working here on my bachelor thesis under supervision of the Net. Sec. Group from ETH Zürich. The goal of the thesis is to analyse the currently implemented beaconing algorithm in terms of bandwidth usage, i.e. analysing its bandwidth overhead and comparing this to possible new strategies.
My intention with this issue is to propose adding bandwidth metrics to the control-service for beaconing and path operations, and announce that I'm working on this actively to avoid duplication of work.
SCION is already using go-grpc-middleware V1 [1] to monitor certain metrics. go-grpc-prometheus [2] is the specific library enabling metric monitoring for gRPC. This library is missing the needed functionality for bandwidth monitoring, but is a good basis to start with. There was already an issue [3] and a PR [4] for integrating bandwidth monitoring to go-grpc-prometheus. The work was put on hold by their maintainers to move now everything to go-grpc-middleware V2.
The idea is to use a different fork from go-grpc-prometheus and integrate the needed functionality to SCION similar to what was already be done in the PR.
[1] https://github.com/grpc-ecosystem/go-grpc-middleware
[2] https://github.com/grpc-ecosystem/go-grpc-prometheus
[3] grpc-ecosystem/go-grpc-prometheus#85
[4] grpc-ecosystem/go-grpc-prometheus#88
The text was updated successfully, but these errors were encountered: