Changelog
Something ships every two weeks.
Every release, newest first, with what it does and why. Subscribe to the status page for incidents; this is the good news.
All releases
Status page subscribers by email and RSS
Customers can subscribe to a status page and get the first update within ninety seconds of an incident opening. RSS for the people who prefer it.
Phone calls read the alert aloud
A page by call now reads the alert's title and the service name, so you know what it is before you find your glasses.
Deploys and rollbacks on the timeline
GitHub and GitLab deploys appear on the incident timeline with the commit, so the postmortem shows what changed and when.
Runbooks show last time's notes
When an alert fires, its runbook opens with what the last three engineers did and how long it took.
Grafana and CloudWatch
Two more sources. Twelve now, plus the webhook.
Escalation by service map
Pages go to the owner in your service map, then up the chain every five minutes. Rotas are optional.