Red Hat Bugzilla – Bug 1014037
[QA] (6.4.0) Clean shutdown of server with deployed JMS bridge hangs
Last modified: 2017-10-09 20:11:15 EDT
Description of problem:
1. Start two EAP 6.x servers - "source" server with deployed queue InQueue and "target" server with deployed queue OutQueue
2. Send 1000 large messages to InQueue
3. Deploy JMS bridge to "source" server which resends messages from InQueue to OutQueue (to "target" server)
4. During processing of messages from InQueue cleanly shutdown "target" server
5. Try to cleanly shutdown "source" server
Clean shutdown of "source" server hangs. Thread dump attached.
"source" server should shutdown.
Created attachment 805812 [details]
I've asked Andy Taylor to take a look on this.
It was just open.. I think it would take us at least 1 week to resolve it. We can do a release if there's space on the release schedule. We will communicate accordingly within a week.
This is still issue in EAP 6.3.0.ER9 (HQ 2.3.20).
Attaching reproducer - reproducer-jms-bridge.zip:
1. Download and unzip reproducer-jms-bridge.zip
2. run "sh prepare.sh"
3. Start target server (server1) - "sh start-server1.sh 127.0.0.1"
4. start source server wiht JMS bridge (which is server3) - "sh start-server3.sh other_ip"
5. Start procucer sending messages to server 3 - "sh start-producer.sh 192.168.40.3 jms/queue/InQueue 1000"
6. When proucer is sending messages cleanly shutdown (ctrl-c) server1
7. Now cleanly shutdown (ctrl-c) server3 -> server hangs
Created attachment 919001 [details]
Im working on this
fixed in https://github.com/hornetq/hornetq/pull/1720
Appears to be fixed by HQ upgrade to 2.3.21 https://bugzilla.redhat.com/show_bug.cgi?id=1132168. Setting to MODIFIED.
This problem appears to be fix. I can't simulate it any-more.
Verified in EAP 6.4.0.DR1.