Repository Docs & Knowledge Base

Consolidated Spensara documentation repository (722 Markdown & architecture documents in ../repository)

Detailed engineering and feature implementation specifications.

implementation-specs/specs/SingleStoreRedshiftRuntimeReadPath/TASK4.1.md 2,596 bytes · Updated Oct 23, 2026 10:44 AM EDT

Raw File

TASK 4.1: Add Smoke Validation for the Runtime Read Path

Add smoke validation proving that the runtime read path can query Redshift curated data loaded by the synthetic feed.

This task validates both migrated runtime consumers:

Add tests or a CLI/dev smoke path that validates:

Exact test project depends on existing project conventions.

Smoke validation must prove:

runtime can resolve the default or explicit tenant,
missing/inactive tenant mappings fail clearly,
data access tools list or expose only approved source-mirror curated objects,
data access tools can query approved KDS source-mirror tables such as curated.kds_orders,
data access tools can query approved KDS source-mirror tables such as curated.kds_items,
non-curated schema references are rejected,
migrated /warehouse/orders endpoints return data from Redshift curated when synthetic data exists.

Recommended validation queries:

Deferred.

The temporary CLI smoke command added during implementation was removed. This task still needs a proper validation path.

[ ] Smoke validation covers tenant resolution
[ ] Smoke validation covers missing or inactive tenant failure
[ ] Smoke validation confirms curated-only schema discovery
[ ] Smoke validation confirms query against approved source-mirror curated objects
[ ] Smoke validation confirms non-curated schema rejection
[ ] Smoke validation covers at least one migrated /warehouse/orders endpoint once that migration exists
[ ] Smoke validation uses Redshift curated data loaded by slice 02
[x] Monitor specs are not migrated in this task
[x] Solution builds without errors

This is smoke validation, not the full monitor evaluation framework.
If local Redshift is not available in normal test runs, keep integration tests clearly gated/configured.
Do not require Athena, S3, COPY, or UNLOAD.