[Research] A clockless vardiff strands a slowing miner

In the current discussion of mining setups involving Stratum-v2 and Datum protocols, the role of a local proxy or datum-gateway is highlighted as a critical component in managing miner efficiency and communication with the pool. This gateway acts as an intermediary that not only controls but also monitors any changes in the mining process—such as a decrease in active miners or a slowdown in hash rates. These variations could be either managed by the proxy/gateway directly via shutdown or slowdown commands or observed through changes like slowed share submissions.

Furthermore, the proxy can effectively aggregate and calculate the collective hashrate of the connected mining hardware by evaluating the variable difficulty (vardiff) of each miner's connection that remains active and has submitted a share within a specified timeframe, typically 30 seconds. The system is designed to adjust the vardiff for each miner to achieve an average share submission interval of approximately three seconds.

On the side of the mining pool, it becomes essential for the pool to adapt its requirements based on the aggregated data received from the proxy. This involves adjusting the pool’s vardiff settings to align with the estimated total hashrate provided by the proxy, ensuring that shares are consistently submitted at regular intervals, around every three seconds in this setup. This interaction between the proxy/gateway and the pool underscores an engineering challenge that focuses on optimizing data exchange and operational efficiency at the datacentre level rather than merely tweaking the vardiff mechanism itself.