CapeCode MQTT accessor leaves threads running and fails to exit

[cxbrooks]

Note: the issue was created automatically with bugzilla2github tool

Original bug ID: BZ#13
From: @cxbrooks
Reported version: unspecified
CC: accessors-devel@terraswarm.org

[cxbrooks]

Running

$PTII/bin/ptinvoke ptolemy.moml.MoMLSimpleApplication ptolemy/actor/lib/jjs/modules/mqtt/demo/MQTTPubSub/MQTTPubSub.xml

fails to exit after the model wraps up.

jstack shows the following threads:

--start--

Full thread dump Java HotSpot(TM) 64-Bit Server VM (25.101-b13 mixed mode):

"Attach Listener" BZ#46 daemon prio=9 os_prio=31 tid=0x00007ff86717a000 nid=0x5a0b waiting on condition [0x0000000000000000]
java.lang.Thread.State: RUNNABLE

"DestroyJavaVM" BZ#27 prio=5 os_prio=31 tid=0x00007ff86721f800 nid=0x1c03 waiting on condition [0x0000000000000000]
java.lang.Thread.State: RUNNABLE

"MQTT Call: mqttpt_FC8DAA38CA21D9F7" BZ#21 prio=1 os_prio=31 tid=0x00007ff8617c0800 nid=0x6703 in Object.wait() [0x0000700009a34000]
java.lang.Thread.State: WAITING (on object monitor)
at java.lang.Object.wait(Native Method)
- waiting on <0x000000076c23c190> (a java.lang.Object)
at java.lang.Object.wait(Object.java:502)
at org.eclipse.paho.client.mqttv3.internal.CommsCallback.run(CommsCallback.java:125)
- locked <0x000000076c23c190> (a java.lang.Object)
at java.lang.Thread.run(Thread.java:745)

"MQTT Snd: mqttpt_FC8DAA38CA21D9F7" BZ#20 prio=1 os_prio=31 tid=0x00007ff8612a4000 nid=0x6503 in Object.wait() [0x0000700009931000]
java.lang.Thread.State: TIMED_WAITING (on object monitor)
at java.lang.Object.wait(Native Method)
- waiting on <0x000000076c23c390> (a java.lang.Object)
at org.eclipse.paho.client.mqttv3.internal.ClientState.get(ClientState.java:572)
- locked <0x000000076c23c390> (a java.lang.Object)
at org.eclipse.paho.client.mqttv3.internal.CommsSender.run(CommsSender.java:96)
at java.lang.Thread.run(Thread.java:745)

"MQTT Call: mqttpt_C7CF5AE1A4F8D261" BZ#17 prio=1 os_prio=31 tid=0x00007ff864800800 nid=0x6103 in Object.wait() [0x000070000972b000]
java.lang.Thread.State: WAITING (on object monitor)
at java.lang.Object.wait(Native Method)
- waiting on <0x0000000779ac73a0> (a java.lang.Object)
at java.lang.Object.wait(Object.java:502)
at org.eclipse.paho.client.mqttv3.internal.CommsCallback.run(CommsCallback.java:125)
- locked <0x0000000779ac73a0> (a java.lang.Object)
at java.lang.Thread.run(Thread.java:745)

"MQTT Snd: mqttpt_C7CF5AE1A4F8D261" BZ#16 prio=1 os_prio=31 tid=0x00007ff864803000 nid=0x5f03 in Object.wait() [0x0000700009628000]
java.lang.Thread.State: TIMED_WAITING (on object monitor)
at java.lang.Object.wait(Native Method)
- waiting on <0x0000000779aa0bf8> (a java.lang.Object)
at org.eclipse.paho.client.mqttv3.internal.ClientState.get(ClientState.java:572)
- locked <0x0000000779aa0bf8> (a java.lang.Object)
at org.eclipse.paho.client.mqttv3.internal.CommsSender.run(CommsSender.java:96)
at java.lang.Thread.run(Thread.java:745)

--end---

To fix this, we need to stop these threads.

Wrapup is calling MQTTHelper.end():

public void end() throws MqttException {
if (_mqttClient.isConnected()) {
_mqttClient.disconnect();
}
_mqttClient.close();
}

https://www.eclipse.org/forums/index.php/t/1073983/ suggests upgrading or calling MqttClient.setTimeToWait().

Upgrading to 1.0.2 from http://central.maven.org/maven2/org/eclipse/paho/org.eclipse.paho.client.mqttv3/1.2.0/ does not solve the problem, though a different set of threads is seen in jstack.

I did not try calling setTimeToWait().