[rmcdaniel]

The parity corpus starts 624 mock Servers. Their default half-second serve_forever poll adds about five minutes of idle shutdown waits to every caller validation. The current exact hosted parity step took 6m31s, so this wait is a concrete bottleneck while shared #297 callers are being migrated.

This supplies a 10ms polling interval to the fixture thread. Request/response handling, corpus, comparisons, timeouts, cleanup and job gates remain unchanged. Ten actual health-request/shutdown cycles measured 5.03s before and 0.12s after, with byte-identical responses and all ports closed. Existing focused parity tests pass; the full hosted corpus and main checks will verify the resulting behavior and timing.

No PHP runtime/profile/catalog or release source changes. This addresses the measured CI wait before queuing the next caller migration, as required by the organization guide.

[sonarqubecloud]

Quality Gate passed

Issues
0 New issues
0 Accepted issues

Measures
0 Security Hotspots
0.0% Coverage on New Code
0.0% Duplication on New Code

See analysis details on SonarQube Cloud

[rmcdaniel]

Security bounded source and CI-throughput review: PASS for exact 26b002c6d30fec2bdf7ccc5b31c7677fd4fdec88 / tree 66799c9754249cbf72015d0899aca86c516ccf1e.

The complete delta from accepted CLI main 5421757314303952359c6a72ce74924bc699112e is one fixture file and one behavioral setting: MockServer supplies poll_interval=0.01 to the existing daemon serve_forever thread. Request handling, routes, parity cases, result comparison, process deadlines and product/release source are unchanged.

The lifecycle remains bounded: run_all creates at most one per-case mock, invokes reference and candidate inside the existing try, and always calls shutdown() plus server_close() in finally. Reducing the internal shutdown poll changes response latency for the stop request, not request semantics. The smaller interval also remains scoped to this short-lived sequential fixture rather than a product server.

Independent exact-head checks:

• python3 -m unittest -v test_parity.py: 25 tests PASS;
• ten real health-request/close cycles: one byte-identical response, 0.129 seconds total, and every bound port refused reconnection after close;
• git diff --check is clean.

This clears the narrow source change. Keep the draft until the two still-running exact-head jobs finish, especially the complete Rust unit/parity job and dry-run verification against the 2.x CLI. No corpus, PHP profile/catalog, caller admission, package or release verdict is inferred.

[rmcdaniel]

Exact-head gate update for the prior Security source verdict: all 28/28 current checks are now terminal SUCCESS at 26b002c6d30fec2bdf7ccc5b31c7677fd4fdec88, including the complete Rust unit/parity job and dry-run verification against the 2.x CLI.

This clears the two pending CI conditions named in the earlier review. The verdict remains scoped to the fixture polling optimization; it does not add corpus, profile/catalog, package or release authority.

[rmcdaniel]

Exact 26b002c6d30fec2bdf7ccc5b31c7677fd4fdec88 has all28 checks terminal SUCCESS. Actual Rust job132272453768 preserves the full corpus:919/947 cases match,28 intentional differences, zero unexpected. Both PHP-vector comparisons, all original Rust checks,63 host unit cases, runtime preparation and dry-run verification pass.

Observed hosted parity-step timings with the admitted PHP profile:

Source	Run/job	Duration
Original3dbda322	43536687894/132267444663	7m12s
Short-poll26b002c	43538238892/132272453768	2m05s

That is5m07s less observed fixture-stage time, with the same947 outcomes. Independent source/timing review 6869107260 and terminal gate 6869123411 pass. Normal merge/main/branch readback follows. No corpus, comparison, timeout, PHP profile/catalog or release source changes.

[rmcdaniel]

Hosted timing readback at the reviewed head: the exact Compare against the 2.x CLI step ran from 01:07:53 to 01:09:58 UTC, 2m05s, versus the recorded 6m31s predecessor. That is a 4m26s reduction, about 68%, while the complete comparison and all 28 checks passed. The full Rust unit/parity job completed in 6m04s.

This provides the requested measured CI-bottleneck evidence for the fixture-only change; the source-scope limits in the prior review remain unchanged.

[rmcdaniel]

Delivered through normal main 59545f90c23a9990b24b97fa5bfdecdc3a58ae70, tree 66799c9754249cbf72015d0899aca86c516ccf1e, exactly equal to reviewed26b002c. Merge-point main equality, remote topic deletion and absence are verified.

All four exact-main workflows are terminal SUCCESS: Build43539150958, CodeQL43539150188, Sonar43539150507, and Public Boundary43539150652.

The reviewed one-file change keeps the same corpus, comparisons, request deadlines and complete cleanup. Fresh hosted proof retains919/947 matches,28 intentional differences and zero unexpected; the comparable active-profile PR stage falls from7m12s to2m05s (exact run/job comparison). The separately noted6m31 predecessor is032's pre-activation PHP setup, a different baseline.

Independent source and terminal gates are 6869107260 and 6869123411. PHP runtime/catalog, package and release authority are unchanged. This removes the measured idle CI wait before the next retained297 caller migration.