The chart can no longer describe a startup budget it does not have

PreWarm__GateReadiness is the portal's safest deploy switch: it holds /health red until the new image's NodeType bake is proven green, so an image that regressed a type stalls its own rollout with the previous image still serving, instead of erroring in front of users.

It is also the switch with the most prerequisites, and the chart lists them carefully — including one that is arithmetic: the startup probe's budget (periodSeconds × failureThreshold) has to cover a full cold bake, measured from production at roughly 2.4 seconds per NodeType, sequential, or about ten minutes on the largest mesh we run. The chart is emphatic about what happens otherwise: Kubernetes kills the pod mid-bake, every time, forever.

And then, two lines later, it told you the budget was already thirty minutes. Twice — once as the "paired setting", once as what "managed envs set". Every environment in the fleet runs five minutes. So the prerequisite was satisfied only in the reader's head, and the first person to arm the gate on that sentence would have walked straight into the failure the sentence was warning them about.

What changed

The comments now describe what the chart actually ships, and say plainly that arming the gate means raising the budget in the same change.

More usefully, the coupling stopped being a paragraph. A new guard fails the build when:

The first half is the one that matters most, because it fails today, on the drifted comment. A guard that only woke up once someone armed the gate would have caught nothing on the day the prose went wrong — which is the whole reason the prose was still there to find.

Reconnecting…
The server was updated. Reloading the page to pick up the latest version.