FireHydrant vs Kubernetes

FireHydrant shipped a real override layer for on-call schedules. Overrides now survive schedule edits made from either the web UI or the API, always win against an overlapping generated shift, can be stacked, and can be cancelled to restore the original system-generated shift. Reassignment, coverage requests, and shift deletion no longer mutate the generated schedule at all — each creates an override instead, which is what makes every change cleanly undoable. The constraint is historical: restoring a shift requires knowing what it originally looked like, and that data has only been tracked since 24 July 2026, so older rotations need a one-time regeneration first.

The last year reads as one project — making an incumbent-grade on-call product safe enough to switch to. Customizable retrospective exports, authenticated status pages, consolidated analytics, an EU instance, MS Teams support, then the Signals Migrator that automates the move off PagerDuty and Opsgenie. Overrides fit the same pattern: a feature teams assume already exists, previously broken badly enough that admins learned not to touch a schedule once coverage was arranged. The work is closing table-stakes gaps ahead of a migration window, not opening new surfaces.

Expect the override model to be pushed out to the API and Terraform surfaces next, along with prompts to regenerate stale rotations before Opsgenie's April 2027 deletion date drives migration volume.