Spike: Research the best way to detect when MySQL is out of temporary space

[jonathansadowski]

There's a suggestion that #634 could be solved if we added a check to see if MySQL is out of temporary space.

Desired outcome of this spike: proposal for a way to detect MySQL has exhausted its temporary space and an understanding of whether or not we could apply that detection toward solving #634.

[barryhughes]

Some relevant docs/articles:

• MySQL Docs ▸ How MySQL Handles a Full Disk
• Cyberciti ▸ MySQL Error 28 and Solution

Possibly we can detect this specific error (whether that's through exception handling, or by inspecting $wpdb->last_error, or some other approach) and then we could perhaps put a temporary pause on further queue processing (as proposed here).

However, we also recently added an extra set of safety checks to prevent the AS-specific problem of actions being processed ahead-of-time. With that in place, I'm wondering if there is still value in doing more research here, or if this can be closed.

cc @jonathansadowski @vedanshujain

[barryhughes]

Closing, since #634 was merged some time back and should avoid the core issue of actions being inadvertently processed ahead of schedule.