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
{{ message }}
This repository has been archived by the owner on Aug 2, 2021. It is now read-only.
Currently a node can be healthy by having neighbours only.
AFAIK, the current idea of healthiness is:
you know of and are connected to all neighbors
if depth value is greater than 0, you should know of and be connected to at least one peer in each bin shallower than depth.
Previously we had a method full which I believe was part of making sure of this. It can be found eg. in 362e2ba
For the abandoned work on the kademlia cleanup I introduced a notion "potent" which was supposed to be this exact check. It also has unit tests. The branch in question is here: #1108
Hopefully the task is not much more than superimposing this work on top of the current head, and making sure that the simulation tests have the right expectations.
The text was updated successfully, but these errors were encountered:
After a call with @zelig it became clear that the work for this ticket is not correctly described.
According to @zelig we need to refactor the notion of Saturated such that the PeerPot is passed the ideal version of the kademlia, and then compare with the real one - that will determine saturation.
The rows below the depth there is a choice of multiple peers to fill the bins, so we compare connections with MinBinSize; the nearest neighbours will need exact match
Currently a node can be healthy by having neighbours only.
AFAIK, the current idea of healthiness is:
Previously we had a method
full
which I believe was part of making sure of this. It can be found eg. in 362e2baFor the abandoned work on the kademlia cleanup I introduced a notion "potent" which was supposed to be this exact check. It also has unit tests. The branch in question is here: #1108
Hopefully the task is not much more than superimposing this work on top of the current head, and making sure that the simulation tests have the right expectations.
The text was updated successfully, but these errors were encountered: