I am having a issue with PI webservices to an MDM Slave repository.
We are receiving a number of warnings about the notification queue, this can cause the Java nodes on PI to go down with out of memory errors.
The scenario's we use are configured with SOAP sender channel sending SOAP requests to PI, We have deployed the MDM services on the PI environment (this is the generated ear file from the MDM repostory) . durring heavy load it seems to trigger this warning [The notification queue contains 9,877 notifications] starting at 1000
Message: Nov 27, 2013 4:19:11 PM com.sap.mdm.internal.protocol.manual.notification.NotificationHandler handleMessage
WARNING: The notification queue contains 9,877 notifications [2013-11-26T16:04:10 2013-11-27T16:19:11 2013-11-27T16:19:11].
Date: 2013-11-27
Time: 16:19:11:794
Category: System.err
Location: System.err
Application: sap.com/WS_MDM_Customer_Query_ENTAPP
Thread: Thread[class com.sap.mdm.internal.notification.InternalEventDispatchercom.sap.mdm.internal.notification.InternalEventDispatcher@4fa4a341,5,Dedicated_Application_Thread]
Data Source: j2ee/cluster/server0/log/defaultTrace_00.trc
Arguments:
DSR Transaction: e3dab59f569011e3c5bd00199991c948
Message Code:
Session: 19
Transaction:
User: MDM_SOAP
Time Zone: +0100
CSN Component:
DC Component: com.sap.mdm.tech.mdm4j
Correlation ID: 201160250000000005
DSR Root Context ID: E3DAB59F569011E3C5BD00199991C948
DSR Connection: e3dab59f569011e3c5bd00199991c948
DSR Counter: 0
Host: cip20
System: P20
Instance: DVEBMGS20
Node: server0
Log ID: 00199991C9481669000022AC000056B5
Question(s)
1) What's the warning about?
2) How can we reset this counter? (java node restart on the PI invironment works but is not an option to do this every day)
Thanks!!