allthingssecurity commented on PR #28637: URL: https://github.com/apache/camel/pull/28637#issuecomment-6489784005
@davsclaus thanks for the review. Changes in 45f45d89704b (one new commit on top): - ExceptionHandler (`CosmosDbConsumer.java:114/128`): the first failed attempt of each start now goes to the consumer's `ExceptionHandler`, as the one failure to start did before; each further attempt is one WARN line. So with `bridgeErrorHandler=true` the error handler gets the failure once per start, not every 30 seconds. A failure that ends after the consumer stopped is not reported, and a throwing handler does not end the retry. New test `theFirstFailedStartGoesToTheExceptionHandlerAndTheFollowingOnesAreLogged` (two failed starts, then a working one: the handler gets only the first); it fails without the change. The 4.24 upgrade guide also has an entry now for the retry and how failures are reported. - `clientWrapper`: `createChangeFeedProcessor()` builds the wrapper as a local, no longer reassigning a field on the `boundedElastic` threads. - reactor-core in the body: I corrected it, but not to 3.8.8. The parent's `reactor-version` (3.8.8) does not manage reactor-core; camel-azure-cosmosdb resolves 3.7.11 from azure-core 1.57.0 (`dependency:build-classpath`). `BlockingSingleSubscriber.blockingGet` checks `Schedulers.isInNonBlockingThread()` first in 3.7.11 and 3.8.7 (javap). The body now says so. camel-azure-cosmosdb: 58 tests, 0 failures (22 ITs skipped); `CosmosDbConsumerFirstStartFailsTest` and `CosmosDbConsumerStopWhileStartingTest` each passed 20 runs in one JVM. Open follow-up: the health check you suggested (the consumer reports started while the processor is not). I would do it as a separate JIRA, a readiness check that is DOWN until the change feed processor has started. _Claude Code on behalf of allthingssecurity_ -- 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]
