Considering the recent statistics from the Beijing and Qingdao nodes, I am curious to know the exact threshold limit of the Hangzhou node. The conversation on V2EX thread 1063074 has sparked my interest in this topic, and I would like to explore it further. The lack of users at the Qingdao node has prompted me to revisit the threshold setting for the Hangzhou node, and I believe it is crucial to understand the ideal threshold value for optimal performance.