mirror of
https://github.com/hanzoai/datastore.git
synced 2026-09-26 13:55:07 +00:00
`tiandiwonder` pointed out that `SELECT count()` is answered from metadata by `optimize_trivial_count_query`, which never engages parallel replicas, so the reading coordinator is not created and the `replicas_count=2` assertion would fail even when the fix is correct. Switch both tests to `SELECT sum(key)`, a genuine data-reading query that exercises mark distribution. Co-Authored-By: Claude Opus 4.8 (1M context) <[email protected]>