Another thing: I was so lucky because get the daily send of digest mails in action. And this line:

if ($digestposts = get_records('forum_queue')) {

just crash if there are too many records in the table (right now one server has more than 800.000!).

I would propose to:

• Change code to use one recordset.
• Add one more query to delete old records say, one week, at the begining of the forum cron. just to avoid current and future accumulation!
• Add some trace message about the digest process starting (was impossible to detect if before!).

And this is the called DIGEST_CRON_PROBLEM from now.

I continue advancing...