I use Multisite and call cron procedures by external (in relation to WordPress) scripts.
Since cron is a table independent for each blog, I implemented a cron call for several sites at the same time. And at night it led to the allocation of all memory and stop all sites.
Reason: every site runs cerber_do_daily. As the table plugin doesn’t depend on the blog, they are usually just multiple (useless) cleaned up and optimized.
However, when you call cron for multiple sites at the same time, the tables are locked and waiting. All accesses to sites also go into waiting for the lock to be released (for writing to tables).
The simplest solution is to call these actions only for one site on the entire network.
So far I’ve just commented out this code (call OPTIMIZE). I hope for a speedy and correct solution of the problem.
This may be related: https://wordpress.org/support/topic/heavy-database-load/
- The topic ‘Table locks when using multisite and simultaneous cron’ is closed to new replies.