We have a cron job. We have alerts to monitor the cron job. We also have a monthly task to check the above are working.

Why? I wrote the first 2 so as far as I'm concerned, I'm going to tick yes on the last.

Maybe for if the cronjob reports false positives and it keeps running while there are errors in log. Manual check to be sure?

Cron job and alerts running on completely different infrastructure?

@donkulator "yes", it's nagios, there's some status stuff that runs on the same hardware that reports to the morhership, the mothership will bitch and moan if it hasn't had any probing goodness in a while

Can't you do another cron job for that?

because there'll always be the one "solution architect" or other that, for some reason, will have access to things and will have a stupid idea about things and therefore WILL fuck up things in ways you could have never imagined.

Whether or not you need manual checks kinda depends on how you built the cron-job

For example: if a cron-job only sends an email "alert! something is wrong" when there's an issue - and you haven't got any such emails for +1 week then you can't be sure if the cron job is functioning or has stopped running.

But if your cron-job sends a "everything was fine this week" you can safely assume it's up and running.