FIX: Node Bad Behaviour Tweaks
Current
"bad" is defined as slashPoints / age, so a new node with 100 points is badder than an old node with 100 points.
The issue is that the network is now very reliable so getting even some slash points can tip a new node over the red line fast (much faster than before), and there is a recent trend of new nodes getting churned out immediately.
Desired
Some things to give new nodes some leeway:
- increase
ObservationDelayFlexibility=10
- increase the grace period for a new node to a full churn cycle period (1-2 days)
- add a min slash points (100 seems fine) to be considered to be bad
Edited by THORChain