Hello,

When provisioning many EC2 instances in a short time span the probing done by Foreman will cause you to run into the AWS rate-limit fairly easily since N threads will all be doing this every ~1 seconds until the instance comes up. For now my solution is to insert a sleep(20) before this line:
https://github.com/theforeman/foreman/blob/develop/app/models/concerns/orchestration/compute.rb#L132

This is mostly because my ruby-fu is awful and I'm under a bit of time pressure so I resorted to a quick 'n dirty fix, I'm not sure what a more permanent and acceptable solution to you guys would be?
Simply increasing the time between probes, or even make it adjustable in some "Don't edit the code of the actual application" way?
Or maybe make it like most DHCP clients do, i.e increase the time between probes or some such?

Oh and I assigned this to EC2 but I imagine this will happen to any service that has a rate-limit on it's API's, I just haven't had the time to test this out yet.

Regards
Johan