Reliability

Minus Two Minutes

Tim Irving: a fix that lands before customers notice shows up in the incident metrics as nothing. Why we pay for heroic recovery and not the two-minute save, from the Key Bridge to Y2K.

· read.zerosevzero.com ↗ · 19m read

Sketch of a container ship approaching a lit suspension bridge at night, drawn in pencil and pink ink

Why we picked it · the editor's summary

Irving's argument starts with a data-model bug: an incident cannot end before it starts, so a problem stopped before it lands is recorded as nothing, or not at all. His example is the Key Bridge, where the mayday came at 1:27 and the collapse at 1:29, and the two minutes were the whole outcome. The same shape recurs in a poisoned deployment caught before rollout or a town evacuated ahead of a lahar: the work that mattered most shows up as a quiet quarter, which then draws scrutiny instead of credit. He asks leaders to treat reliability as the disaster that never arrived, not as recovery speed. Our own dashboards would fail his test. We are not sure what we would replace them with, and the piece does not pretend to know either.

Read on zerosevzero.com ↗

Comments

Loading comments…

The Friday deploy, in your inbox

The week’s best writing on shipping software, picked by people who ship.

We’ll email you a confirmation first. Your address is used only to send you this newsletter, and every issue has an unsubscribe link. Privacy policy