fix: offload sync Pydantic serialization to thread in async write path

[charles-dyfis-net]

AsyncPointsApi methods for upsert_points, batch_update, and update_vectors call _build_for_* which runs jsonable_encoder (thus Pydantic's model_dump_json) synchronously before the first await. This blocks the event loop for the entire duration of Pydantic serialization, which scales with point count and vector dimensionality.

Move the sync _build_for_* call into run_in_executor for these three methods (skipping some others where body size is bounded and small). The _build_for_* methods return an unawaited coroutine object (from the async api_client.request), so the executor runs the serialization in a thread and we await the resulting coroutine back on the event loop.

Fixes #1175

[netlify]

❌ Deploy Preview for poetic-froyo-8baba7 failed.

Name	Link
🔨 Latest commit	478165e
🔍 Latest deploy log	https://app.netlify.com/projects/poetic-froyo-8baba7/deploys/69c5786f7ba1c60008a3eed1

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 9c7b6c8c-248d-4f6f-8bae-f2bf16ce747b

📥 Commits

Reviewing files that changed from the base of the PR and between b76b850 and 478165e.

📒 Files selected for processing (1)

• qdrant_client/http/api/points_api.py

🚧 Files skipped from review as they are similar to previous changes (1)

• qdrant_client/http/api/points_api.py

---

📝 Walkthrough

Walkthrough

Added import asyncio and updated three async methods in AsyncPointsApi — batch_update, update_vectors, and upsert_points — to offload synchronous _build_for_* calls to a background thread using asyncio.to_thread(...), capturing the returned coroutine and awaiting it. SyncPointsApi methods were not changed.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~12 minutes

Suggested reviewers

• tbung

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title accurately describes the main change: offloading synchronous Pydantic serialization to a thread in the async write path.
Description check	✅ Passed	The description is clearly related to the changeset, explaining the problem (sync serialization blocking event loop), the solution (moving build_for* calls to executor/thread), and linking to issue #1175.
Linked Issues check	✅ Passed	The PR meets the core objective from #1175: offloading CPU-bound Pydantic serialization to a thread for three async write methods (upsert_points, batch_update, update_vectors) to prevent event loop blocking.
Out of Scope Changes check	✅ Passed	All changes are scoped to the three target async methods; no unrelated changes were introduced beyond the intended thread offloading implementation.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[charles-dyfis-net]

Some contribution process notes:

• Troubleshooting was AI-assisted: I presented Claude with a stack trace from production performance troubleshooting tools -- quoted from in Pydantic serialization invoked by async client can block main loop #1175 -- and interactively walked/talked through the relevant code with the model.
• I did not run the pre-commit hooks because they're tied specifically to Python 3.10, and I don't keep an interpreter that old handy. I have run the pytest suite, and the only error was in unrelated code due to a version mismatch (my local qdrant server is somewhat behind).
• A Netlify error installing qdrant_sphinx_theme in the "deploy preview" does not appear to be relevant to the changes at hand.

[charles-dyfis-net]

Going to revise this slightly to use asyncio.to_thread() in place of run_in_executor.