CONTROL PLANE • REPLAY EXECUTIONS
Re-run incidents with pinned inputs — and prove parity before production.
Replay Executions let teams reproduce incident conditions on a captured telemetry window using a specific policy version. You get a parity report, diffs, and an auditable run record — so automation is validated before it impacts production.
Replay executions create an auditable record of what was evaluated, what changed, and what would happen before production actions run.
WHAT IS A REPLAY EXECUTION?
A tracked run that reproduces incident formation.
A replay execution re-processes a captured telemetry window through TraceFlux using a pinned correlation policy version. It produces artifacts (parity report, diffs, evidence bundle) and records governance (who triggered it, approvals, outcomes).
Replay the same 2–10 minute incident window with consistent inputs.
Compare results across policy changes without guessing.
Approval gates + audit logs for every execution and artifact.
EXECUTION LIFECYCLE
A repeatable pipeline from capture → parity → promotion.
Run replays to validate incident formation, tune policies, and prove automation safety before enforcing changes.
Select an incident window and scope (service, region, POP).
Choose mode: Dry-run vs Enforced. Pin policy version and controls.
Re-process telemetry through correlation and incident aggregation.
Inspect parity report + diffs. Approve policy changes or actions.
RECENT EXECUTIONS
Run history that reads like a real control plane.
Enterprise operators expect an auditable record of job runs and outcomes — replay executions are first-class events in TraceFlux.
Validate changes without triggering production actions. Treat enforcement as a promoted state.
Scope executions to a region/POP/service and require approvals before widening impact.
Parity reports and diffs attach directly to the incident narrative and audit record.
ARTIFACTS
Outputs you can share, audit, and defend.
Replay executions are only valuable if they produce durable, reviewable artifacts — not just “it worked on my machine.”
What changed vs baseline and why. Confidence + scope included.
Fingerprint, suppression, and correlation diffs across policy versions.
Signals + timeline + metadata packaged for review and tickets.
Who triggered, who approved, what gates passed, and what ran.
NEXT STEP
See replay executions on your telemetry.
We’ll walk through capture → correlation → incident formation → replay validation and show how parity reports and governance gates prevent risky automation.