Deadlock due to message queue overrun

When I stress the device with one class 1 I/O connection using a cifX-Scanner and rapidly open class 3 connections in rapid succession and execute Achilles test suites "IP Unicast Storm, IP Multicast Storm, IP Broadcast storm, IP Fragmented storm" in parallel, then the message queue of the object task gets flooded with Abort-Transport-Indications which stem from the cyclic receive timeout event.

This was tracked down to the cyclic auto-reloading TLR timer "hRxTimer" used by the encapsulation task. Auto-reloading timers have worse effects in overload-scenarios in general and should be avoided.