Skip to content

Build your service catalog

Your catalog is the backbone of Reloqui — evidence is attributed to services, and everything else (releases, 360 views, lineage) hangs off it.

Products group related services for rollup views. Catalog → Products → New. Set a name, slug, description and owner team/email.

Catalog → Services → New. Fields that matter:

  • Repository URL — the attribution key. Evidence whose repo matches this URL is attributed to the service. Set it precisely (e.g. https://github.com/acme/checkout-api).
  • Kindservice (deployable), component (a shared library), or resource (database, queue, bucket).
  • Tier & lifecycle — severity/importance and active/archived.
  • Owner team / email — surfaces on Service 360 and in ownership filters.
  • Linksrunbook, dashboard, adr URLs.
  • Custom fields — up to 20 scalar keys (e.g. oncall, sloTarget, dataClass).

Set kind to component or resource. They appear under Catalog → Components / Resources and can participate in the dependency graph, but aren’t release-scoped the way services are.

Catalog → Environments — create development, staging, production, etc. Mark production requiresApproval. Environments drive the per-environment rollout view derived from deployment evidence.

Declare typed edges between services to unlock lineage and blast-radius analysis:

  • calls — runtime dependency
  • depends_on — build/library dependency
  • publishes_to — event/topic relationship

Add edges in Catalog → Dependency graph, or in bulk via the API:

POST /api/v1/catalog/relationships/batch
{
"projectId": "proj_…",
"items": [
{ "fromServiceId": "svc_web", "toServiceId": "svc_checkout", "relType": "calls" }
]
}

Then explore the interactive map at Catalog → Dependency graph, or query lineage:

GET /api/v1/services/{serviceId}/lineage

The lineage response includes upstream/downstream services, recent artifacts and deployments, and shared dependencies, plus a blast radius count.

Reloqui flags catalog gaps (missing repo URL, owner, product) in Catalog → Data quality. Work the queue to keep attribution accurate.

  • Filter lists by ownerTeam, lifecycle, tier, kind, productId.
  • For GitHub, use Discover repos on the connector to suggest services from your org’s repositories.