When the algo runs, it will at every decision record the current local HR to keep track of what it sent where, to be able to calculate the external HR based on sent and total.
When using Proxy, algo should watch his HR instead of xmrig, but doesn't and record only HR of the latter when sending all HR to a node(p2pool or XvB), in case of no shares or 24h average fast reach.
No consequence until a change of decision that will stop sending all HR to a node. Then too few sent HR will be estimated which could over raise the estimated external HR.
Having an overestimate of external HR can lower the required HR to be sent. The algo should resolve the issue with next decisions up to an hour when it realize external HR doesn't exist in the total HR, except if a new decision direct all HR to a node and the behavior repeats again.
Version
Since Gupaxx v1.3.0
Bug
When the algo runs, it will at every decision record the current local HR to keep track of what it sent where, to be able to calculate the external HR based on sent and total.
When using Proxy, algo should watch his HR instead of xmrig, but doesn't and record only HR of the latter when sending all HR to a node(p2pool or XvB), in case of no shares or 24h average fast reach.
No consequence until a change of decision that will stop sending all HR to a node. Then too few sent HR will be estimated which could over raise the estimated external HR.
Having an overestimate of external HR can lower the required HR to be sent. The algo should resolve the issue with next decisions up to an hour when it realize external HR doesn't exist in the total HR, except if a new decision direct all HR to a node and the behavior repeats again.
This bug doesn't not impact HR on XvB because of https://github.com/Cyrix126/gupaxx/issues/33
Steps
Steps to reproduce the behavior