How to fix InnoDB: page_cleaner: 1000ms intended loop took XXXXms?

Question

MySQL

On the server, logs often contain similar errors:
InnoDB: page_cleaner: 1000ms intended loop took 48659ms. The settings might not be optimal. (flushed=236 and evicted=0, during the time.)

The settings are as follows:
innodb-buffer-pool-instances = 32
innodb-page-cleaners = 32
innodb-lru-scan-depth = 256

Not many requests at this point , but the base starts to slow down at this moment. Tell me how you can play around with the settings or in which direction to dig?

Replication is configured from the server. If that plays a role.