We are having a similar issue. CPU/IOWAIT spikes every 30 minutes, causing queries to time out obtaining locks, breaking the application. iotop confirms the MySQL process is the culprit. Enabling SQL logs:

SET GLOBAL general_log = 'ON';

did not reveal either any intensive queries nor a flood of queries.

What other logs can we look at that would identify what MySQL is doing to cause these i/o spikes?

We're running:

mysql  Ver 15.1 Distrib 10.3.36-MariaDB, for Linux (x86_64) using readline 5.1