Help users debug issues when AWS could not provision instances

[deliahu]

Description

Sometimes the auto scaling group fails to spin up instances, e.g. if the user is at their instance limit (spot or on demand), or if AWS is out of capacity (less common but has happened, especially for spot instances). It should be more obvious to the user that this has happened. Usually this information is contained in the Activity History for the auto scaling group

Update: This is now covered for issues that arise during cortex cluster up, but not covered after the cluster has spun up (e.g. if cortex deploy stalls because the spot instance could not be provisioned). This could be handled with another input signal into the API status.

[soulbliss]

Check this page, for more information. On this page and more specifically here.

Updating here as the link mentioned in the above gist has been removed and shifted to the docs.