[scale-infra] incident response

Hi folks,

We're increasingly seeing outages on the website this year. So far we've gotten by on KC or I noticing and rebooting things. While we obviously need to stop the outages from happening at all, we probably should still have some plan in place for how we respond to them - especially during critical times of the year (eg today for reg, sponsorship, and CFP).

I know this isn't anyone's "job", but given how critical the infra is to the conference, should we setup an oncall rotation for it? We seem to be getting caught at various times where nobody with access is available to respond.

Thoughts?

Ilan