Wildly 10 - Embedded Artemis - Dead Client Ressource cleanup problems
vincent.sourin Mar 14, 2016 9:06 AMHello,
We've got a c# application (using Apache NMS) consuming messages as a durable consumer from a Topic.
It works well except when / if the c# application crashed without cleaning its ressources.
When the c# application crashes, we can see in wildfly logs that Artemis detects the dead client and tries to clean up its ressources but seems to failed (AMQ221021 : failed to remove connection)
[Server:Test-Nono] 13:23:28,898 WARN [org.apache.activemq.artemis.core.server] (activemq-failure-check-thread) AMQ222067: Connection failure has been detected: AMQ119014: Did not receive data from /10.10.0.69:53826. It is likely the client has exited or crashed without closing its connection, or the network between the server and client has failed. You also might have configured connection-ttl and client-failure-check-period incorrectly. Please check user manual for more information. The connection will now be closed. [code=CONNECTION_TIMEDOUT] [Server:Test-Nono] 13:23:28,899 WARN [org.apache.activemq.artemis.core.server] (activemq-failure-check-thread) AMQ222061: Client connection failed, clearing up resources for session ID:IT-5-53825-635935585472996610-1:0:-1 [Server:Test-Nono] 13:23:28,902 WARN [org.apache.activemq.artemis.core.server] (activemq-failure-check-thread) AMQ222107: Cleared up resources for session ID:IT-5-53825-635935585472996610-1:0:-1 [Server:Test-Nono] 13:23:28,912 WARN [org.apache.activemq.artemis.core.server] (activemq-failure-check-thread) AMQ222061: Client connection failed, clearing up resources for session ID:IT-5-53825-635935585472996610-1:0:1 [Server:Test-Nono] 13:23:28,913 WARN [org.apache.activemq.artemis.core.server] (activemq-failure-check-thread) AMQ222107: Cleared up resources for session ID:IT-5-53825-635935585472996610-1:0:1 [Server:Test-Nono] 13:23:28,915 INFO [org.apache.activemq.artemis.core.server] (activemq-failure-check-thread) AMQ221021: failed to remove connection
Because of that (or at least we think it is because of that), it's impossible for the c# client to reconnect to the broker. We receive each time this kind of error :
Client: XXXXXXX already connected from /10.10.0.69:53826
The only way we found to be able to reconnect is to restart Wildfly server but it is not a really helpful workaround ;-).
Do we miss something or do we hit a bug ?
Regards,
Vincent