fix(longRunningMigrations): make 0020 batch size small to avoid OOM

[noliveleger]

💭 Notes

• Comment-only change in 0020_backfill_asset_version_hash.py.
• Replaces the previous comment with a clearer explanation of why CHUNK_SIZE is kept small: each version_content blob can be large, so a small batch size trades throughput for a lower memory footprint.
• No behavior change.

[coderabbitai]

Important

Review skipped

Auto reviews are limited based on label configuration.

🏷️ Required labels (at least one) (1)

• llm-ignore

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro

Run ID: 318a8094-cf89-4581-b756-a71729746c4d

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Use the checkbox below for a quick retry:

• 🔍 Trigger review

---

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

[greptile-apps]

Greptile Summary

This PR updates the comment on CHUNK_SIZE in 0020_backfill_asset_version_hash.py to better explain the memory-vs-throughput trade-off, and removes the now-unused settings import. The new comment is clearer and more descriptive.\n\nKey observations:\n- The new comment is a genuine improvement: it names the specific field (version_content JSON blobs) and the explicit goal (lower memory footprint).\n- CHUNK_SIZE is changed from a dynamic min(settings.LONG_RUNNING_MIGRATION_SMALL_BATCH_SIZE, 50) to a hardcoded 5, and the django.conf.settings import is correctly removed as a result.\n- The only open style suggestion is to optionally mention the order-of-magnitude memory impact in the comment itself, to help future maintainers calibrate the value.\n- PR template / title note: The description does not include the ### 📣 Summary, ### 📖 Description, ### 👷 Description for instance maintainers, or ### 🗒️ Checklist sections from .github/PULL_REQUEST_TEMPLATE.md. For an internal-only change those top sections may legitimately be omitted, but the ### 👷 Description for instance maintainers section would be the natural place to describe the performance impact of this constant change on live instances. The PR title also omits the DEV-XXXX ticket number called for by the template checklist (item 4).

Confidence Score: 4/5

Safe to merge; the only open items are a style suggestion on the comment wording and a PR-template compliance note.

The code change itself is minimal and correct — the comment is clearer, the hardcoded CHUNK_SIZE = 5 is intentional, and the dead import is removed. No logic paths are altered. A prior review thread flagged the behavioral aspect of the CHUNK_SIZE reduction; that concern stands but doesn't block correctness. The small deduction from 5 is for the open PR-template compliance gaps (missing maintainer description section, no ticket number) and the style-level opportunity to quantify the memory impact in the comment.

No files require special attention; the single changed file is straightforward.

Important Files Changed

Filename	Overview
kobo/apps/long_running_migrations/jobs/0020_backfill_asset_version_hash.py	Comment updated to explain memory-vs-throughput trade-off; CHUNK_SIZE hardcoded to 5 and unused settings import removed. Logic is otherwise unchanged.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A([run]) --> B{get_queryset\nreturns records?}
    B -- "No records left" --> Z([Done])
    B -- "Up to CHUNK_SIZE=5\nrecords" --> C[Iterate with\niterator chunk_size=5]
    C --> D[Compute content_hash\nfor each AssetVersion]
    D --> E[bulk_update\n_content_hash field]
    E --> F[sleep 2s\navoid DB flood]
    F --> B

Loading

_{Reviews (2): Last reviewed commit: "docs(long_running_migrations): improve C..." | Re-trigger Greptile}