I'm getting a lot of stale share rejects mining cryptonight. It can mine clean for hours but then it will start rejecting
20-25% (up to 50% on 5 min average) of the submitted shares. This occurs with both the Nicehash CPU miner as well
as cpuminer-opt with the Nicehash mod. It also occurs on all miners at the same time.
It seems to be clearly a pool issue. Is it being investigated? It may not hurt buyers very much but it's a significant
penalty for miners.
Where are you from? Note, that we do not have HK and TK proxies up for cryptonight yet, you are being directed directly to US server.
We also checked some miners (top ones with highest speed) and there are some miners with high stale issues. Since there are only some and there is no correlation between them and because some miners are completely clean (especially big ones), we assume this must be a networking issue between us and you. Also, a stale share happens when miner is "too late", which also suggests a networking issue.
What you may try is to mine on different server, EU/US.
Thank's for the response. I switched one of my miners from usa to eu and will monitor.
A networking issue could explain why all miners are affected simultaneously, however, only the cryptonight algo is affected
which suggests something specific to that algo.
I will update when the next wave of rejects occurs.
Edit: I just got a cluster 5 rejects from EU while USA stayed clean. When all miners were on USA the waves of rejects were
always in sync.