[NIFI-7722] Enhance error-handling during request replication

• Type: Bug
• Status: Resolved
• Priority: Major
• Resolution: Fixed
• Fix Version/s: 1.12.0
• Component/s: Core Framework
• Assignee: Tamas Palfy
• Reporter: Tamas Palfy
• Created: 10/Aug/20 17:10
• Updated: 12/Aug/20 16:39
• Resolved: 12/Aug/20 16:39

Description

In ThreadPoolRequestReplicator.NodeHttpRequest.run() the main logic is in a try block, during which all Exceptions are caught and handled properly. However we need to call an Object.notify() after that so it's crucial that the error handling covers every scenario.

Because of this we should catch all Throwables (instead of only Exceptions), otherwise - should we encounter an Error for example - we might end up in a deadlock state.

Issue Links

• links to: GitHub Pull Request #4466