## What changes were proposed in this pull request? User aborted task reported as FAILED in background operations: 1. Install cluster with all services ( Including Druid) 2. Click on Stop all services 3. When BGoperations window is loaded, click on the abort button for the same task 4. The operation should be aborted and marked as Aborted with "Aborted by user" message in the sub tasks logs 5. But instead the operation is marked as failure.
## How was this patch tested? In addition to JUnit tests I used a little [Python script](https://issues.apache.org/jira/secure/attachment/12941947/abort_bg_operations.py) to reproduce the issue first (see in the attachment of the corresponding JIRA). Moreover I edited `ambari-agent.ini` and reduced command report/heartbeat periods. This was needed due to the nature of this issue. When executed my tool (should abort 500 `start_all` - or `stop_all` - requests successfully) it failed very early (90% of the cases did not survive until the 30th round). After I deployed the fix and executed the tool it succeeded. [ Full content available at: https://github.com/apache/ambari/pull/2411 ] This message was relayed via gitbox.apache.org for [email protected]
