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
The first idea which comes to mind is to emit peer score snapshot (immutable) every decayInterval. The snapshot should include both the sore values (including all topics) and original counters (like meshDeliveries, firstDeliveries) for better context. The client may want to subscribe to these snapshots and log some explanation whenever the score drops below some threshold(s)
However there are cases when the score should be updated earlier than at the end of decayInterval, for example invalidMessages when we need to react faster to avoid DoSing with invalid messages for a longer period.
Open questions
No response
Are you planning to do it yourself in a pull request ?
Maybe
The text was updated successfully, but these errors were encountered:
Description
Would be good to have a mechanism to track and explain peer gossip score
Motivation
A peer may loose its mesh due to downscore and we basically have no idea what's going on with a peer score.
https://hackmd.io/fumYP6OKTLSin6dAVSI-9A
Requirements
The first idea which comes to mind is to emit peer score snapshot (immutable) every
decayInterval
. The snapshot should include both the sore values (including all topics) and original counters (likemeshDeliveries
,firstDeliveries
) for better context. The client may want to subscribe to these snapshots and log some explanation whenever the score drops below some threshold(s)However there are cases when the score should be updated earlier than at the end of
decayInterval
, for exampleinvalidMessages
when we need to react faster to avoid DoSing with invalid messages for a longer period.Open questions
No response
Are you planning to do it yourself in a pull request ?
Maybe
The text was updated successfully, but these errors were encountered: