The latest version of SOFARegistry, through some optimizations, currently supports the service discovery of tens of millions of data volume clusters. Compared with the old version, the overall resource cost has also been greatly reduced, but there are still some optimization points that need to be further improved:
Due to the large amount of fixed-delay batch processing, the push delay is still high. It is hoped that by identifying the amount of data, the fixed delay of batch processing can be reduced, and the overall change push delay can be reduced.

Hope to improve the timeliness of pub/sub push;