A status page maintained since 2022, updated eleven times, none of them during an actual outage.
why: updating it was step 9 of the incident runbook,
after the technical steps, and by step 9 the incident
was over.
what moved:
step 1 acknowledge
step 2 POST the status page: investigating
step 3 everything else
and a command that does step 2 in one line, so it is
not a browser tab during an incident.Putting the communication before the diagnosis feels wrong and is right — the people reading the status page need to know somebody is looking, not what is wrong. Making it a command rather than a web form is what got it done, because a browser login at three in the morning is a step everybody skips.