Boost MySQL Connection Performance with Thread Cache: How It Works & Tuning Tips

MySQL improves the performance of the “client request to create a connection” process by using a Thread Cache pool that stores idle connection threads instead of destroying them after each request.

When a new connection request arrives, MySQL first checks the Thread Cache pool; if an idle thread is available it is reused, otherwise a new thread is created.

Related parameters

thread_cache_size : the maximum number of idle connection threads that the Thread Cache can hold. Threads are added to the cache as they are created and used, and once the cache reaches this size MySQL stops adding more idle threads.

thread_stack : the amount of memory allocated for each connection thread’s stack when it is created. The default value (192 KB) is suitable for most environments, and the size should not be changed unless you are very familiar with MySQL’s thread handling.

To monitor the cache, view the related status variables (see the image below).

Thread Cache hit rate is the ratio of connections obtained from the cache to the total number of connections received by the system:

Threads_Cache_Hit = (Connections - Threads_created) / Connections * 100%

After the system has been running stably for a while, the Thread Cache hit rate should generally stay around 90 %.