Set performance bounds for CI and local

[k-raina]

Summary

• Update performance bounds for local and CI environment
• Enhanced Async Producer Strategy (tests/ducktape/producer_strategy.py), added polling logic to async producer to match sync producer behavior

JIRA : https://confluentinc.atlassian.net/browse/NONJAVACLI-3971

[confluent-cla-assistant]

🎉 All Contributor License Agreements have been signed. Ready to merge.
_{Please push an empty commit if you would like to re-run the checks to verify CLA status for all contributors.}

[Copilot]

Pull Request Overview

This PR updates performance bounds for both local and CI environments and enhances the Async Producer Strategy with polling logic to match sync producer behavior.

• Updated performance bounds with more relaxed latency thresholds for both local and CI environments
• Added polling logic to the async producer strategy with configurable interval and metrics tracking

Reviewed Changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 2 comments.

File	Description
tests/ducktape/producer_strategy.py	Added periodic polling logic to async producer with metrics tracking
tests/ducktape/producer_benchmark_bounds.json	Increased p95 and p99 latency thresholds for local and CI environments

---

_{Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.}

[MSeal]

Should we also change the tests to go back to failing test execution on bounds failure to block PRs?

[k-raina]

@MSeal Thanks! FIxed

[sonarqube-confluent]

Analysis Details

1 Issue

• 0 Bugs
• 0 Vulnerabilities
• 1 Code Smell

Coverage and Duplications

• No coverage information (66.90% Estimated after merge)
• No duplication information (4.90% Estimated after merge)

Project ID: confluent-kafka-python

View in SonarQube

[fangnx]

Do you think we should just remove min_messages_per_poll, as messages_per_poll is essentially equal to the batch_size according to our code