📢 Incident Report: Hivemind Service Disruption - January 26

The website started to experience unstable access from 21:30 on January 26th, Beijing time.

The general situation is as follows: The execution of complex queries in the hive_posts_cache table led to a sharp increase in database I/O.

The database response became slower, and finally, the connection pool between Hivemind and the database was filled up, causing the service to be unresponsive.

At present, a previous TODO regarding caching has been completed, increasing the utilization rate of the cache.

In addition, the capacity of the database connection pool has been increased.

Let's observe further to see if the problem will occur again.