Define GitHost Emergency SLA

With our improvements to monitoring and such, I think we need to consider defining a GitHost Emergency SLA or some goal.

Ticket which spurred the idea: https://gitlab.zendesk.com/agent/tickets/65354

I have had prospective customers have ask about this. If they have EE Starter we commit to a 24 hour response time. Based on the fact that we do monitoring I'd guess we will take it upon ourselves to fix things. But as it stands it appears we advertise 24 hour response to support items -- including this. Clearly that is not the case. I'm not sure where this needs to be documented but we do need to document it!

mentioned in issue #80

My instance is down and the monitoring clearly doesn't work. I don't really think a response time of 24 hours when the instance is down (especially not because the upgrade happened on a unscheduled time) is acceptable. Second time this happens in the middle of the day within a week.

@lbot in the context of this ticket: https://gitlab.zendesk.com/agent/tickets/80062 - do you think this makes sense to re-visit?

My thinking is that if we are the admins of the instance, we should offer a level of support equal to that of an internal admin at the customer.

cc @pidge