rohankumardubey opened a new pull request, #4269: URL: https://github.com/apache/iggy/pull/4269
## Which issue does this PR address? Closes #3941 ## Rationale Five consecutive NACKs could permanently stop a source after a short broker outage. For sources holding accepted input only in memory, stopping can lead to data loss, and the runtime did not reliably report that the poll task had ended. ## What changed? Source implementations can now provide their own batch-result timeout and consecutive NACK limit. The HTTP source disables the NACK limit by default for accepted webhooks, while other sources retain the existing limit. A source that can safely replay a rejected batch can also opt to keep polling after a transient NACK with capped backoff. The runtime is notified when a source poll task stops unexpectedly, updates its status and running gauge, and clears a stale error after a successful send. SDK documentation and regression tests cover these paths. ## Local Execution - Passed: `cargo fmt --all`, `cargo sort --no-format --workspace`, and workspace Clippy with `-D warnings`. - Passed: SDK, connectors runtime, HTTP source, and PostgreSQL source tests; relevant binaries/plugins built. - Passed: targeted Markdown lint for the updated SDK README. - Not passed: the integration harness could not reach connector tests because the local server mesh failed to start. - Pre-commit hooks did not run; the relevant checks above were run manually. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
