---
title: "Replay all persisted active or failed rules onto RTVI VLM"
method: POST
path: "/api/v1/realtime/replay"
tags: ["realtime"]
---

# Replay all persisted active or failed rules onto RTVI VLM

`POST /api/v1/realtime/replay`

Trigger Alert Bridge to re-onboard every persisted active or failed rule from Elasticsearch onto RTVI VLM. Intended for use after RTVI VLM restarts. Returns 409 if a replay is already running. While a replay is in progress, POST and DELETE on /realtime return 503.

## Response `200`

Replay completed

- RealtimeReplayResponse — Response for POST /api/v1/realtime/replay.
  - `status` string
  - `message` string, required
  - `correlation_id` string, required — Per-invocation UUID4 hex woven through every replay log line (stage=replay_start|replay_rule|replay_end). Operators grep their log aggregator with it to follow one invocation end-to-end across the per-rule fan-out.
  - `replayed` integer, required — Number of rules successfully re-onboarded
  - `failed` integer, required — Number of rules that failed re-onboard
  - `total` integer, required — Total rules attempted
  - `details` object[] — Per-rule outcome: id, alert_type, result, error (if failed)

## Other responses

- `409` — Replay already in progress
- `501` — Persistence not configured
- `502` — Elasticsearch read failure

---

[API](https://skmtc.dev/nvidia/apis/alert-agent-api.md) · [All operations](https://skmtc.dev/nvidia/apis/alert-agent-api/llms.txt) · [OpenAPI document](https://skmtc-service-production.skmtc.workers.dev/v1/apis/nvidia/alert-agent-api/revisions/e754ab3efedd/schema)