Driftwire 4.2: scheduled rollouts and automatic rollback
Rollouts that advance on a schedule while every guardrail holds, and roll themselves back the moment one does not. Plus stale-flag reports and a faster CLI.

Driftwire 4.2 is out. The headline is that a rollout no longer needs someone watching it: tell it where to go and what must stay true on the way, and it gets there on its own — or comes back.
Scheduled rollouts
Define the steps once. Driftwire advances to the next step at the time you set, but only if every guardrail has held for the whole of the current one.
// What an edge node receives: a diff against the version it holds.
interface RulesetDiff {
version: number
previous: number
changed: Record<string, FlagRules>
removed: string[]
}
function apply(diff: RulesetDiff) {
if (diff.previous !== store.version) return resync() // missed one: fetch a snapshot
store.merge(diff.changed, diff.removed)
store.version = diff.version
}Automatic rollback
If a guardrail breaches for two minutes, traffic goes back to 0%, the flag owner is paged, and the audit log records which metric tripped it. Nothing to remember at 2 a.m.
Also in 4.2
- Stale-flag reports: flags at 0% or 100% for 30 days are flagged for removal, with the code references to delete.
- The CLI starts four times faster:
driftwire flags listreturns in 90 ms. - SSO group sync for any SCIM provider.
The full list is in the changelog. Every plan gets 4.2 today.

Written by
Priya Raman
Product Lead, Releases
Priya leads product for releases and experiments. Previously ran the deploy pipeline for a payments company where a bad rollout was a bad day for a lot of people.
More posts by Priya Raman
