There is unexpected server shutdown during in test scenario - "OOM in remote JCA topology in cluster with restart"
1. Start EAP servers 1 and 3 in HornetQ cluster (jms cluster) with deployed queues InQueue and OutQueue
2. Start other 2 EAP servers 2 and 4 (mdb servers) which has configured resource adapter to connect to jms cluster
3. Send 50 000 small to InQueue to jms cluster
4. Deploy MDB to mdb servers
- MDB consumes messages from InQueue and for each message send new message to OutQueue. MDB makes JNDI lookup for OutQueue for each message.
5. During processing of messages by MDB cause out of memory on heap on server 2
6. Wait 5 minutes and restart server 2
7. Wait some time until the rest (if there are some) of messages are consumed from InQueue
8. Receive messages from OutQueue
Test failed in step 6. when "Critical IO Error, shutting down the server." was logged on server-3 and server shuted down. There are set lower values for max-size-bytes (1MB) and max-size-bytes (100KB) to stress paging.
Customer impact:
if client is disconnected in the moment when new page file is being created. It leads to unexpected interrupt of page file creation which as consequence shutdowns the server.