Host provisioning jobs with scopes cause job backups if unable to finish successfully

I created a spawn host with a volume that doesn't exist in AWS since it was available in the spawn host UI (it's probably a bug to allow people to try attaching volumes that don't exist in AWS) and it got stuck in provisioning (couldn't finish the user data done job because attachVolume would always fail on the AWS call). This seems to have caused amboy to refuse to run other jobs in the queue.

Splunk errors: https://mongodb.splunkcloud.com/en-US/app/search/search?q=search%20index%3Devergreen%20i-0b3a0f488d280099a&display.page.search.mode=smart&dispatch.sample_ratio=1&earliest=1586550600&latest=1586561220&sid=1586561901.485818