Basis Errors and Solutions

Symptom

You notice that, the J2EE Engine goes into a never ending cycle of restarts, due to OutOfMemory Error. This error is noticed in the thread dumps and default traces as follows:
“FATAL: Caught OutOfMemoryError! Node will exit with exit code 666″
=======================================================================
java.lang.OutOfMemoryError

Other terms

SAP XI AF Messaging ,SAP XI AF ,OOM ,Out Of Memory ,restart ,J2EE , server restart
Reason and Prerequisites

Huge messages are capable of causing an OutOfMemoryError,especially on the receiver side. Each time, when it is picked up for processing, it crashes the engine, leading to an endless loop of restarts.
Solution

The endless loop of restarts is broken by “blacklisting” of messages. In this feature, following actions are performed on a server restart:
a) the set of messages that were persisted with state “DLNG” in the database, would be flagged as ‘blacklist candidates’.The blacklist candidates would then be changed to “TBDL” state.
b) the set of messages that were already flagged as ‘blacklist candidates’, during the previous restart, would be considered as ‘bad messages’. These bad messages would then be changed to “NDLV” state, (rather than “TBDL” state), to ensure that, they are not picked up for processing again. Consequently, the problems of bad messages are alleviated, by limiting the number of crashes(to two, rather than an endless loop) they can cause. Blacklisted messages in state “NDLV” can only be restarted by explicit action of an Administrator.
In order to activate this behaviour, the parameter “messaging.blacklisting.enabled” must be set to “true” for the “SAP XI AF Messaging” service in the Visual Administrator.
The archives and the support package stack guide can be found on the SAP Service Marketplace as described in SAP Note 952402.