Element metadata in windowed value for offset deduplication

[stankiewicz]

Currently offset deduplication is hidden in KafkaIO by forcing redistribute.
If user wants to do any kind of parsing, filtering or data size reduction, they can't benefit from it.

This changes:

• creates elementMetadata within WindowedValue, there are plans to add more fields in upcoming future
• plumbs ProcessContext, DoFn to make it possible to propagate elementMetadata, this will be simplified soon with Introduce OutputBuilder in Java SDK #34902
• allow user to skip redistribute at their own risk
• propagates offset details in WindowedValue metadata up to WindmillSink

Limitations:

• element metadata is not shuffled
• offset deduplication works only with Dataflow runner v1

@kennknowles

---

Thank you for your contribution! Follow this checklist to help us incorporate your contribution quickly and easily:

• Mention the appropriate issue in your description (for example: addresses #123), if applicable. This will automatically add a link to the pull request in the issue. If you would like the issue to automatically close on merging the pull request, comment fixes #<ISSUE NUMBER> instead.
• Update CHANGES.md with noteworthy changes.
• If this contribution is large, please file an Apache Individual Contributor License Agreement.

See the Contributor Guide for more tips on how to make review process smoother.

To check the build health, please visit https://github.com/apache/beam/blob/master/.test-infra/BUILD_STATUS.md

GitHub Actions Tests Status (on master branch)

See CI.md for more information about GitHub Actions CI or the workflows README to see a list of phrases to trigger workflows.

[gemini-code-assist]

Summary of Changes

Hello @stankiewicz, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

I've implemented a significant enhancement to how metadata is handled within Beam's data processing model, particularly focusing on KafkaIO. The core idea is to expose element-specific metadata, such as Kafka offsets, directly within the WindowedValue structure. This change allows users to perform operations like parsing or filtering on Kafka records before offset-based deduplication, even if they choose to bypass the redistribution step. This provides greater flexibility and efficiency for certain data processing patterns, especially for Dataflow runner v1. The changes involve introducing a new ElementMetadata class, plumbing this metadata through the DoFn and ProcessContext interfaces, and updating KafkaIO to attach relevant offset details, which are then consumed by the WindmillSink for deduplication.

Highlights

• New ElementMetadata Class: I've introduced a new ElementMetadata class to encapsulate record-specific details like ID and offset, which can now be attached to WindowedValue instances.
• WindowedValue Enhancement: The WindowedValue structure has been updated to include this new ElementMetadata, allowing for richer context to be carried alongside data elements.
• Metadata Propagation: I've plumbed the ElementMetadata through DoFn and ProcessContext interfaces and their implementations across various runners and SDK components, ensuring this metadata is accessible throughout the pipeline execution.
• KafkaIO Integration for Offset Deduplication: For KafkaIO, I've added the capability to extract Kafka record offsets and topic/partition information and attach them as ElementMetadata to the WindowedValue. This enables offset-based deduplication to function even when redistribution is skipped.
• WindmillSink Utilization: The WindmillSink in the Dataflow runner has been updated to consume and utilize this new ElementMetadata for its offset-based deduplication logic, leveraging the propagated Kafka offset details.
• Flexible KafkaIO Deduplication: Users can now opt to skip the withRedistribute() step in KafkaIO while still benefiting from offset deduplication, though a warning is logged about potential data loss if elements are mutated or aggregated without redistribution.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point in your pull request via creating an issue comment (i.e. comment on the pull request page) using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands.

Feature	Command	Description
Code Review	/gemini review	Performs a code review for the current pull request in its current state.
Pull Request Summary	/gemini summary	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in issue comments and review comments.
Help	/gemini help	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments or fill out our survey to provide feedback.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Footnotes

1. Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

[gemini-code-assist]

Code Review

This pull request introduces ElementMetadata to WindowedValue to support offset-based deduplication in KafkaIO without forcing redistribution. The changes are extensive, touching many core components to plumb the new metadata through the system. Overall, the approach is sound, but I've identified a few critical issues, including potential null pointer exceptions and an incomplete method implementation that could lead to data loss. I've also noted some opportunities to reduce code duplication and improve an error message for better maintainability and clarity.

[github-actions]

Assigning reviewers:

R: @m-trieu for label java.
R: @fozzie15 for label kafka.

Note: If you would like to opt out of this review, comment assign to next reviewer.

Available commands:

• stop reviewer notifications - opt out of the automated review tooling
• remind me after tests pass - tag the comment author after tests pass
• waiting on author - shift the attention set back to the author (any comment or push by the author will return the attention set to the reviewers)

The PR bot will only process comments in the main thread (not review comments).

[github-actions]

Reminder, please take a look at this pr: @m-trieu @fozzie15

[stankiewicz]

r: @kennknowles

[stankiewicz]

r: @sjvanrossum for kafka

[github-actions]

Stopping reviewer notifications for this pull request: review requested by someone other than the bot, ceding control. If you'd like to restart, comment assign set of reviewers

[kennknowles]

When I do finish the OutputBuilder PR, I think that ElementMetadata will be deleted. So it would make my life simpler if you added the two fields in that class instead of adding the extra layer of indirection object.

[stankiewicz]

Run Java PreCommit

[stankiewicz]

Run Java_GCP_IO_Direct PreCommit

[kennknowles]

LGTM, thanks for kicking the tests until they are green

[stankiewicz]

thanks, could we merge it?

[kennknowles]

Just checking - this is not preserved through the windowed value coder yet.

[stankiewicz]

That is correct, this is something I'm working now for v1