A bad experience in Heroku

So the story starts with two things in Heroku and Rails world:

1. Delayed Job - https://github.com/collectiveidea/delayed_job/
2. One off process - https://devcenter.heroku.com/articles/oneoff-admin-ps

I have used delayed job for event and other mail notifications. and tested it by runing the following rake locally;

rake jobs:work

All went fine, got mail notifications and all.

Now, I deployed to application to Heroku for production setup. and mistakenly added rake jobs:work to the scheduler and set it to run in every hour.

What it does internally for every run it starts a one-off process of Heroku which means starting one instance. As per one-off process, it has to be close manually or the program that triggers it should close. BUT

rake jobs:work; does not close its on going process, so every run it started one one-off process, and finally at the end of one month, there were hell amount of one-off processes, and hence shoot up the bill unexpectedly.

Thanks to my manager; who didn't charged that from me, A Nice Gentleman.

But I was thinking, there should be an alarm if this kind of expected event happen in Heroku.

Tip of the day:

Please follow exactly what Heroku says in it Document.

Al the best!

--
Regards,
Pavan