High CPU load due to RCF AMI thread

[Volker]

Hi Jarl,

we're observing that one of the threads in our process starts occupying one core of our CPU. This happens after stopping the process hosting the server. As our clients mainly need to become informed about new events from the server the clients instantiate a server on their own to establish a callback connection. After destroying this 'callback server' in the client process the thread starts his phase of high load. Interestingly this has no influence on the functionality of our application.

As this thread calls itself "RCF AMI" my first bet is that RCF started this thread. We tried to extract something valuable from the callstack of this thread but without success. Our workaround is that we no longer destroy the 'callback server' but reuse it.

Regards,

Volker

[jarl]

Hi Volker,

I've had a look at this, and found a shutdown problem with a timer on the AMI thread. I've fixed that and will email you a copy of the source code to verify whether it fixes the problem you've observed.