[NIFI-6697] HandleHttpRequest - 500, stopping does not works

• Type: Bug
• Status: Resolved
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 1.8.0
• Fix Version/s: 1.10.0
• Component/s: Core Framework
• Environment: CentOS Linux release 7.5.1804 (Core) ; Oracle jre1.8.0_181-amd64
• Assignee: Unassigned
• Reporter: Mermillod
• Created: 20/Sep/19 16:59
• Updated: 17/Nov/21 14:22
• Resolved: 17/Nov/21 14:22

Description

My HandleHTTPRequest does not accept anymore POST messages, after some time, returning 500 (port 10010).

There is lot of file availaible on the machine (ulimit etc).

Settings :

• 1 thread
• timerdriver
• 200ms
• 50 containerqueue (i'm pretty sure i never fill it due to low volumes of calls).

"Stop Processor" does not work : it starts the stop thread, never ending.

"Terminate Thread" seems to work (the #2 on the processor disappears), but a thread dump shows that not :  

/"qtp920046370-777080-acceptor-0@367cdcb6-ServerConnector@6962170f{HTTP/1.1,[http/1.1]}{0.0.0.0:10010}" Id=777080 RUNNABLE
 at sun.nio.ch.ServerSocketChannelImpl.accept0(Native Method)
 at sun.nio.ch.ServerSocketChannelImpl.accept(Unknown Source)

waiting on java.lang.Object@3b97aac8 at sun.nio.ch.ServerSocketChannelImpl.accept(Unknown Source)
waiting on java.lang.Object@3b97aac8 at org.eclipse.jetty.server.ServerConnector.accept(ServerConnector.java:369) at org.eclipse.jetty.server.AbstractConnector$Acceptor.run(AbstractConnector.java:639) at org.eclipse.jetty.util.thread.QueuedThreadPool.runJob(QueuedThreadPool.java:762) at org.eclipse.jetty.util.thread.QueuedThreadPool$2.run(QueuedThreadPool.java:680) at java.lang.Thread.run(Unknown Source)

 

The listener is still there : 

 

netstat -an | grep 10010
 tcp 0 0 0.0.0.0:10010 0.0.0.0:* LISTEN 

 

Restarting the processor (after the "terminate thread") does not work since the port is already listening.

 

Current workaround : stop/start nifi.

 

Issue Links

• is fixed by: NIFI-6317 HandleHttpRequest timeout handling issues