Skip to documentation content
Dataset APIDocs

Troubleshooting

Diagnose common Dataset API integration failures from observable evidence.

In this guide
  1. 1Start with request ID
  2. 2Verify stable IDs
  3. 3Check release quality
  4. 4Reconcile builds
In this guide
  1. 1Start with request ID
  2. 2Verify stable IDs
  3. 3Check release quality
  4. 4Reconcile builds

Start with the smallest reproducible request

  1. 1
    Call GET /health with the same credential and environment
  2. 2
    Call GET /datasets and confirm the dataset is visible
  3. 3
    Capture status, request_id, route, and timestamp
  4. 4
    Reduce the failing query to release_id and limit=1
  5. 5
    Compare the request with the canonical OpenAPI operation

Common symptoms

SymptomLikely causeNext check
401 on every routeMissing, wrong-environment, expired, or revoked keyAuthorization header and key rotation state
404 for a releaseWrong dataset/release pair or unavailable releaseList releases for the dataset
Empty query resultExact filter does not match or wrong release selectedRemove filters and request one row
partial=trueSelected release has incomplete quality statusInspect the release manifest and policy
Build remains queuedCapacity or upstream dependency delayBuild age, health, and operational deadline
Duplicate buildsPOST retried after ambiguous timeoutList builds and reconcile recent requests

Schema mismatch after a release change

Fetch the dataset definition and release manifest again. Compare dataset_version, definition_version, schema_version, required fields, and field types with the last successful import. Unknown optional fields should not fail tolerant readers, but a missing required field must stop the import visibly.

Collect safe support context

Was this page helpful?