Ops Limits
Ingest volume, query windows, concurrent jobs, and retention.
Limits are per Ops project. Silent truncation does not happen. Caps return structured errors.
Incidents and jobs
| Limit | Enterprise default |
|---|---|
| Open incidents | 500 |
| Summaries per incident per hour | 20 |
| Concurrent summaries | 15 |
| Timeline events per incident | 10,000 |
| Post-mortem revises | 6 |
Logs
| Limit | Enterprise default |
|---|---|
| Query window | 24 hours (SEV-1 keys can request 72h) |
max_lines | 20,000 |
| Scan lines per query | 5,000,000 |
| Concurrent log queries | 8 |
| RPM log query | 60 |
409 query_window_too_large if to - from exceeds the cap. Narrow the window.
Ingest (OTel)
| Limit | Enterprise default |
|---|---|
| Spans per minute | 200,000 |
| Log records per minute | 500,000 |
| Connector lag alert | 120 seconds |
Ingest overage drops with ops.connector.lag and a 429 on the ingest host, not on api.geteridian.com.
Catalog
| Limit | Enterprise default |
|---|---|
| Services | 2,000 |
| SLOs per service | 8 |
| Connectors | 25 |
Retention
| Record | Default |
|---|---|
| Incident and timeline | 365 days |
| Log query results | 30 days (ids remain on the incident) |
| Post-mortems | 365 days (or legal hold via Risk if linked) |
| ZDR | GET of summary bodies 404 after completion |
Follow-the-sun
Human support for the Eridian control plane: Dubai, Hong Kong, South Africa. These limits do not change on weekends. SEV-1 on your services is your rotation. SEV-1 on Eridian: contact@geteridian.com with X-Eridian-Request-Id.
See Ops API and Rate Limits.
Production API credentials are issued with an institution workspace. Contact sales if you need access.