[bokelley]

Unblocks CI on next. Since 2026-11-01, Postgres conformance tests (Postgres 16, core) has failed on every PR (#1764, #1765) with:

FAILED tests/conformance/reporting/test_reporting_row_storage_object_pg.py::test_the_producer_stages_object_uploads_outside_its_publication_transaction
ReportingRowStoreError: Reporting row store: STATE_UNAVAILABLE: no live worker lease covers this row upload

Cause

The test runs one producer turn with the settling harness, whose clock is fixed at 2026-11-01T02:10Z. The PG ledger store stamps the worker lease from that injected clock. The row upload (adcp_reporting_storage/ledger/row_write.py) then checks that lease against the database's clock_timestamp(). That passed while the real time was before the harness instant plus the lease. It fails on every run since the wall clock passed it. The last green next run was on 2026-10-31.

Fix

The test sets the harness clock to the wall clock before the turn, so the lease lies in the real future. The test still checks what it checked before:

• the upload is staged outside the publication transaction;
• one object is written;
• rows read back, and verified rows read back;
• no write intent is left.

Production code is unchanged.

Reproduced and verified against a local PostgreSQL 16: the test failed before the change, and the whole file passes after it (23 passed).

🤖 Generated with Claude Code

[aao-secretariat]

Ladon verdict: Approve

Approve — clean test-only change.

Verified: single-file diff (7 net lines) in tests/conformance/reporting/test_reporting_row_storage_object_pg.py:443-457. The change captures the harness clock and advances clock[0] = datetime.now(timezone.utc) after setup and before run_worker(), so the stamped worker lease lands in the real future relative to the DB clock_timestamp() check — fixing the STATE_UNAVAILABLE failure that appeared once the wall clock passed the fixed harness instant. Production code, assertions, wire shape, public surface, generated types, and governance are all untouched.

No critical/high/medium findings. gated_paths is false, high_risk is false, no no-auto-approve team match, no prior decision. None of rows 1–8 fire; falls through to row 9.