Chapter 9. Detecting Dead Connections
Sometimes clients stop unexpectedly and do not have a chance to clean up their resources. If this occurs, it can leave resources in a faulty state and result in the broker running out of memory or other system resources. The broker detects that a client’s connection was not properly shut down at garbage collection time. The connection is then closed and a message similar to the one below is written to the log. The log captures the exact line of code where the client session was instantiated. This enables you to identify the error and correct it.
[Finalizer] 20:14:43,244 WARNING [org.apache.activemq.artemis.core.client.impl.DelegatingSession] I'm closing a JMS Conection you left open. Please make sure you close all connections explicitly before let
ting them go out of scope!
[Finalizer] 20:14:43,244 WARNING [org.apache.activemq.artemis.core.client.impl.DelegatingSession] The session you didn't close was created here:
java.lang.Exception
at org.apache.activemq.artemis.core.client.impl.DelegatingSession.<init>(DelegatingSession.java:83)
at org.acme.yourproject.YourClass (YourClass.java:666) 1- 1
- The line in the client code where the connection was instantiated.
9.1. Connection Time-To-Live
Because the network connection between the client and the server can fail and then come back online, allowing a client to reconnect, AMQ Broker waits to clean up inactive server-side resources. This wait period is called a time-to-live (TTL). The default TTL for a network-based connection is 60000 milliseconds (1 minute). The default TTL on an in-VM connection is -1, which means the broker never times out the connection on the broker side.
Configuring Time-To-Live on the Broker
If you do not want clients to specify their own connection TTL, you can set a global value on the broker side. This can be done by specifying the connection-ttl-override element in the broker configuration.
The logic to check connections for TTL violations runs periodically on the broker, as determined by the connection-ttl-check-interval element.
Procedure
Edit
<broker_instance_dir>/etc/broker.xmlby adding theconnection-ttl-overrideconfiguration element and providing a value for the time-to-live, as in the example below.<configuration> <core> ... <connection-ttl-override>30000</connection-ttl-override> 1 <connection-ttl-check-interval>1000</connection-ttl-check-interval> 2 ... </core> </configuration>
9.2. Disabling Asynchronous Connection Execution
Most packets received on the broker side are executed on the remoting thread. These packets represent short-running operations and are always executed on the remoting thread for performance reasons. However, some packet types are executed using a thread pool instead of the remoting thread, which adds a little network latency.
The packet types that use the thread pool are implemented within the Java classes listed below. The classes are all found in the package org.apache.actiinvemq.artemis.core.protocol.core.impl.wireformat.
- RollbackMessage
- SessionCloseMessage
- SessionCommitMessage
- SessionXACommitMessage
- SessionXAPrepareMessage
- SessionXARollbackMessage
Procedure
To disable asynchronous connection execution, add the
async-connection-execution-enabledconfiguration element to<broker_instance_dir>/etc/broker.xmland set it tofalse, as in the example below. The default value istrue.<configuration> <core> ... <async-connection-execution-enabled>false</async-connection-execution-enabled> ... </core> </configuration>
Additional resources
- To learn how to configure the AMQ Core Protocol JMS client to detect dead connections, see Detecting dead connections in the AMQ Core Protocol JMS documentation.
- To learn how to configure a connection time-to-live in the AMQ Core Protocol JMS client, see Configuring time-to-live in the AMQ Core Protocol JMS documentation.
