[tedzhouhk]

Note

Stacked PR, live retune (split from #17562). Merge order: #17483 → #17733 (this PR) → #17734 → #17735 → #17736 → #17562

This PR targets hzhou/dgdr-trace-pvc, the branch of #17483 (the PR below it), so its diff shows only this layer. #17562 was split into these layers for review. At the split they carried #17562's original diff exactly, plus a 40-line replay check in #17734 that rejects executor: dgdr in offline replay; the review follow-up of 2026-11-02 has since changed several layers.

Overview

Live planner retune (#17562) compares Candidates that different DGDR Runs of one window scored. That comparison only works if a Candidate's score depends on its own configuration and nothing else. This PR makes DynamoReplayRunner replay every trace workload with common random numbers (CRN): ordinal request ids and the canonical router seed. DynamoReplayRunner is the dynamo replay runner that AISimulate's Sweeper loads (entry point in pyproject.toml), and it is what a DGDR trace Run executes.

run_trace_replay already has the internal _deterministic option for this; offline replay retune uses it for its baseline and candidates. This PR only passes it from the Sweeper runner. The synthetic-workload path is unchanged.

This layer changes no file that #17483 touches. It sits on #17483 for merge order, because DGDR trace Runs (spec.workload.trace) come from #17483.

Key Changes

• components/src/dynamo/replay/simulation.py: passes _deterministic=True to both run_trace_replay calls of DynamoReplayRunner (agg and disagg). run_synthetic_trace_replay has no such parameter and does not receive it.
• components/src/dynamo/replay/api.py: comments only. _deterministic is no longer described as internal to retune sweeps: every trace replay through the Sweeper's dynamo runner sets it.
• components/src/dynamo/replay/tests/test_simulation.py: the existing test_trace_runner_preserves_current_replay_arguments now runs for an agg and a disagg deployment, checks each one's worker counts, and asserts _deterministic is True.

Behavior note: this applies to every trace replay through the Sweeper's dynamo runner, including every DGDR trace Run, not only the Planner's. A Run's scores no longer depend on the other candidates or on the Run's parallelism.

Validation

Ran on this layer (e9ab6afc24, before the review follow-up) and compared with the same suites at the stack base 6e8087d359 (#17483's head, with main c6b055d4c7 merged in):

Check	Result
replay/tests/test_simulation.py::test_trace_runner_preserves_current_replay_arguments (agg, disagg)	2 passed, run at the top of the stack (92cee14c88), where simulation.py and test_simulation.py are identical to this layer's head 8e2df8215e. The rest of the file: 54 passed, and 2 failed on this host's missing conversation_affinity (see below).
Planner and replay suites (components/src/dynamo/planner/tests, components/src/dynamo/replay/tests, no xdist)	At e9ab6afc24: 59 failed, 2137 passed, 2 skipped. At the base: 59 failed, 2131 passed, 2 skipped. The two sets of failing tests are identical; the 6 extra passes were the tests of the earlier separate CRN test file, which the follow-up folded into test_simulation.py. Not re-run per layer after the follow-up.
pre-commit run --files on the changed files	At e9ab6afc24, on its 2 changed files: all hooks pass, no files changed. After the follow-up, pre-commit from 8d098bed48 to 92cee14c88 (the top of the stack, covering every layer's follow-up) is clean.
CI mypy (the dynamo-pipeline.yml command, MYPYPATH=components/src:lib/bindings/python/src mypy --explicit-package-bases) on dynamo/planner and dynamo/replay	At e9ab6afc24: 9 errors, the same 9 as at the base: missing yaml/pandas/nvtx stubs, plus existing arg-type errors in common/http/url_validator.py and planner/simulation/provider.py. None is new.

The 59 failures come from this host. Its prebuilt bindings lack conversation_affinity (#16854) and the ais-forward-pass feature, and it has no etcd binary. CI builds the bindings from source.

The development-cluster run in #17562 used a planner image with this change for its Run Jobs, on #17562's pre-split branch. Both windows of that run reached a decision (kept, applied). The executor allows a decision only when the CRN probe's score matches the grid evaluation of the same point within a relative error of 1e-9. This layer was not run on a cluster by itself.

Issues and Ongoing Work

• Part of live retune (feat(planner): enable live retune through DGDR v1beta2 #17562), the retune work in feat(planner): engine-owned retune with replay router control #17095 and the runtime reconfiguration DEP DEP (light): Planner runtime reconfiguration with predictor warmup #16835.
• The next layer is feat(planner): add the live retune configuration and deployment premises #17734.

Review follow-up (2026-11-02)

• Folded the determinism check into the existing test_trace_runner_preserves_current_replay_arguments, with an agg and a disagg row, and removed the separate CRN test file.
• Cut the runner comment down to the invariant: trace replays use ordinal request ids and the canonical router seed, so a candidate's score depends only on its configuration and the trace.
• Updated the replay/api.py comments: _deterministic now applies to every trace replay through the Sweeper's dynamo runner, not only to retune sweeps.
• Merged feat(planner): live retune from recorded traces through DGDR v1beta2 #17483's review follow-up into this branch.

Pull Request Checklist

• I used Dynamo's /pr-checklist skill.

🤖 Generated with Claude Code

[devin-ai-integration]

✅ Dynamo PR CI passed — run 41524032893 (attempt 1) on ce59903e10

Gate checks: ✅ backend-status-check · ✅ deploy-status-check · ✅ dynamo-status-check

_{Posted automatically by Devin for run 41524032893. Updated on every full-CI run of this PR.}

[jthomson04]

Reviewed at e9ab6af. The change is fine: the canonical seed is the same for every candidate and window, no user seed is overridden, and the offline-only guard can't trip. Two nits inline.