tju-yxq opened a new issue, #2659: URL: https://github.com/apache/rocketmq-dashboard/issues/2659
## Problem `POST /api/instances/delete-batch` is documented and implemented as a per-instance operation that returns partial results. The loop currently catches only `BusinessException`, but Apache provider preflight and cleanup can also throw a runtime failure when the broker or local dependency is unavailable. If the first selected instance fails that way, the exception leaves `InstanceService.deleteInstances(...)` immediately: - the later selected instances are never attempted; - the endpoint returns a generic 500 instead of the batch result; - the caller cannot tell which items succeeded or failed. ## Expected behavior A failure while processing one selected identifier should be recorded for that identifier and must not prevent the remaining identifiers from being attempted. Existing `BusinessException` messages and the successful deletion count should remain unchanged. Unexpected failures should be reported with a bounded, readable message rather than an internal stack trace. ## Verification Add a service regression test where the first Apache instance's `countTopics` throws `IllegalStateException` and the second instance deletes successfully. The test should prove both instances are attempted, the result contains one deletion and one failure, and no exception escapes the batch endpoint. -- 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]
