I wanted to report that this issue is still very much alive a year after being reported here. A
'*/5 * * * *' cron schedule for an action varies wildly from a five minute tick. Just in the last 24 hours I’ve seen the 5 minutely job run anywhere from 6 to 45 minutes between runs.
While the job should be running 12 times per hour, I don’t think I’ve ever seen it run more than 5 or 6 times in a one hour period.
If the system is unable to operate reliably, you should at least put a warning in the
schedule docs to indicate that running on schedule is more of an aspiration, and should not be expected in practice.
You seem to imply this is a scaling issue, which if true has been present for a very long time. Strangely though, if I manually trigger an action via UI or a push it always runs within seconds of the trigger. So this clearly seems to be an issue only with cron triggers.
I hope you can raise this issue again.