currently, I found in some cases (e.g. the database instance is stuck, so all query responses will be affected), the PMM agent uses too many connections, e.g.
+----------+----------------+---------------------+
| count(*) | user | ts |
+----------+----------------+---------------------+
| 306 | pmm | 2022-10-20 10:06:59 |
| 306 | pmm | 2022-10-20 10:07:07 |
| 306 | pmm | 2022-10-20 10:07:15 |
| 306 | pmm | 2022-10-20 10:07:23 |
the PMM agent issues more than 300 connections to the host it monitors, How to limit the number of connections PMM uses in case the PMM query response is delayed? It’s indeed a problem, this will make the server status even worse.