Solved Higher temporal resolution for at(1)?

at(1) jobs are only processed every 5 minutes, and the man page notes this nicely:
IMPLEMENTATION NOTES
Note that at is implemented through the cron(8) daemon by calling
atrun(8) every five minutes. This implies that the granularity of at
might not be optimal for every deployment. If a finer granularity is
needed, the system crontab at /etc/crontab needs to be changed.
So changing it sounds relatively straightforward, except when I run $ grep atrun /etc/crontab, I see nothing. It's not in the system crontab, or any user crontab that I can see, yet my at jobs seem to still run (albeit, only at 5-minute intervals). So, what is the proper way to get per-minute resolution for at?
 
Ah, so it's in a separate system crontab in cron.d. Out of curiosity, why are some things in the system crontab (e.g. periodic) and others broken out to their individual files in cron.d? There seems to be no rhyme or reason as to what is where.
 
The cron.d directory is a relatively new thing, and the “atrun” job has been moved to it recently. I assume that this is an ongoing process, so more jobs will be separated out in the future.

That makes sense, because it makes it easier to have local changes. For example, I always add the -a OLD -t DEFAULT options to the newsyslog line. Therefore I’m looking forward to that line being separated out, too, so I won't have to change the global /etc/crontab file anymore.
 
Back
Top