Summary
Read-only endpoint that exposes streaming-pipeline metrics for a warehouse connection. Powers the lightweight telemetry widget on the Manage flow (FE-B) and the detailed Streaming Status page (Screen 2, FE-D).
Endpoint
GET /api/v1/organisations/{organisation_pk}/integrations/warehouse-connections/{id}/telemetry/
Request body
None.
Response (200 OK)
{
"status": "active",
"last_delivery_at": "2026-04-20T11:00:00Z",
"events_last_24h": {
"evaluations": 148233,
"custom_events": 2104,
"total": 150337
},
"events_last_7d": {
"evaluations": 1022104,
"custom_events": 18207,
"total": 1040311
},
"failed_deliveries_last_24h": 0,
"recent_errors": [
{ "occurred_at": "2026-04-19T23:14:00Z", "message": "..." }
]
}
status enum
active — delivering normally
degraded — some failures in the last 24h but the pipeline is catching up
error — not delivering (dead-letter queue growing or all batches failing)
idle — no events received
Field usage
- FE-B (Manage flow card) reads
status, last_delivery_at, events_last_24h.total.
- Screen 2 (Streaming Status page) reads everything, including per-type breakdowns and
recent_errors.
Notes
- Separate from
GET /warehouse-connections/{id}/ so the main resource call stays cheap and cacheable.
- Counts aggregated from the streaming pipeline's internal metrics (delivered via Workstream 2).
- No side effects — pure read.
- Returns
404 if the connection doesn't exist.
- Returns a "no data yet" shape (zeroes, nulls,
status: idle) if the pipeline hasn't streamed anything yet.
Summary
Read-only endpoint that exposes streaming-pipeline metrics for a warehouse connection. Powers the lightweight telemetry widget on the Manage flow (FE-B) and the detailed Streaming Status page (Screen 2, FE-D).
Endpoint
Request body
None.
Response (
200 OK){ "status": "active", "last_delivery_at": "2026-04-20T11:00:00Z", "events_last_24h": { "evaluations": 148233, "custom_events": 2104, "total": 150337 }, "events_last_7d": { "evaluations": 1022104, "custom_events": 18207, "total": 1040311 }, "failed_deliveries_last_24h": 0, "recent_errors": [ { "occurred_at": "2026-04-19T23:14:00Z", "message": "..." } ] }statusenumactive— delivering normallydegraded— some failures in the last 24h but the pipeline is catching uperror— not delivering (dead-letter queue growing or all batches failing)idle— no events receivedField usage
status,last_delivery_at,events_last_24h.total.recent_errors.Notes
GET /warehouse-connections/{id}/so the main resource call stays cheap and cacheable.404if the connection doesn't exist.status: idle) if the pipeline hasn't streamed anything yet.