OpenFire: can't understand missing replies

Hi all,

I'm experiencing a strange behavior in OpenFire 3.9.3 that I'm not able to reproduce systematically: I've a Window service that connects multiple clients (built on MatriX), let's say 1 every 5 seconds. It occasionally happens that I get no OnLogin/OnBind events and so I quit for timeout after some time.

If I look at the client logs, nothing happens until my timer expires; if I look at the server, nothing is reported in Error/Warn/Info logs, but I found two different patterns in Debug:

when everything goes fine I see:

2015.10.14 17:33:20 org.apache.mina.filter.executor.ExecutorFilter - Launching thread for /0:0:0:0:0:0:0:1:64881

2015.10.14 17:33:20 org.apache.mina.filter.executor.ExecutorFilter - Exiting since queue is empty for /0:0:0:0:0:0:0:1:64881

2015.10.14 17:33:20 org.apache.mina.filter.executor.ExecutorFilter - Launching thread for /0:0:0:0:0:0:0:1:64881

2015.10.14 17:33:20 org.apache.mina.filter.executor.ExecutorFilter - Exiting since queue is empty for /0:0:0:0:0:0:0:1:64881

2015.10.14 17:33:20 org.apache.mina.filter.executor.ExecutorFilter - Launching thread for /0:0:0:0:0:0:0:1:64881

2015.10.14 17:33:20 org.jivesoftware.openfire.nio.NIOConnection - NIOConnection: startTLS: using c2s

2015.10.14 17:33:20 org.jivesoftware.openfire.nio.ClientConnectionHandler - [/0:0:0:0:0:0:0:1:64881]  doHandshake()

2015.10.14 17:33:20 org.jivesoftware.openfire.nio.ClientConnectionHandler - [/0:0:0:0:0:0:0:1:64881]   handshakeStatus=NEED_UNWRAP

2015.10.14 17:33:20 org.jivesoftware.openfire.nio.ClientConnectionHandler - [/0:0:0:0:0:0:0:1:64881]  unwrapHandshake()

2015.10.14 17:33:20 org.jivesoftware.openfire.nio.ClientConnectionHandler - [/0:0:0:0:0:0:0:1:64881]    inNetBuffer: java.nio.DirectByteBuffer[pos=0 lim=0 cap=16921]

2015.10.14 17:33:20 org.jivesoftware.openfire.nio.ClientConnectionHandler - [/0:0:0:0:0:0:0:1:64881]    appBuffer: java.nio.DirectByteBuffer[pos=0 lim=33842 cap=33842]

2015.10.14 17:33:20 org.jivesoftware.openfire.nio.ClientConnectionHandler - [/0:0:0:0:0:0:0:1:64881]  Unwrap res:Status = BUFFER_UNDERFLOW HandshakeStatus = NEED_UNWRAP

bytesConsumed = 0 bytesProduced = 0

2015.10.14 17:33:20 org.apache.mina.filter.executor.ExecutorFilter - Exiting since queue is empty for /0:0:0:0:0:0:0:1:64881

and when it doesn't I see:

2015.10.14 17:33:40 org.apache.mina.filter.executor.ExecutorFilter - Launching thread for /0:0:0:0:0:0:0:1:64890

2015.10.14 17:33:40 org.apache.mina.filter.executor.ExecutorFilter - Exiting since queue is empty for /0:0:0:0:0:0:0:1:64890

2015.10.14 17:33:40 org.apache.mina.filter.executor.ExecutorFilter - Launching thread for /0:0:0:0:0:0:0:1:64890

2015.10.14 17:33:40 org.apache.mina.filter.executor.ExecutorFilter - Exiting since queue is empty for /0:0:0:0:0:0:0:1:64890

2015.10.14 17:33:50 org.apache.mina.filter.executor.ExecutorFilter - Launching thread for /0:0:0:0:0:0:0:1:64890

2015.10.14 17:33:50 org.apache.mina.filter.executor.ExecutorFilter - Exiting since queue is empty for /0:0:0:0:0:0:0:1:64890

I'm not sure about that, but it seems to me it just says that the thread exits without doing the doHandshake() stuff.

Someone can give me a hint?

Thank you for your replies