[NIFI-6567] HandleHttpRequest does not shutdown HTTP server in some circumstances

• Type: Bug
• Status: Resolved
• Priority: Minor
• Resolution: Fixed
• Component/s: Extensions
• Assignee: Peter Wicks
• Reporter: Peter Wicks
• Created: 19/Aug/19 16:04
• Updated: 11/Sep/19 15:29
• Resolved: 11/Sep/19 15:29

Description

(Dependent on NIFI-6562)

If multiple cluster nodes are running on a single host, then there will be port conflicts in HandleHttpRequest.  To avoid this users may choose to run in Primary Only scheduling mode.

If there is a primary node change, the processor does not properly shutdown the old Http server instance before the new primary node tries to startup it's own local copy.

Issue Links

• links to: GitHub Pull Request #3673