00030010: EOC due to DEREGISTER of partner.

I am working with EntireX using RPC server running on UNIX. I tested Natural part separately using samll test Natural program simulating external call and everything works fine, no errors. Then I created a test client in .Net and started calling Natural Server program from there.
And now I am having this DEREGISTER message and can not get rid of it.
Going through the program step by step I can see that it goes ok about half way through and then dying.
Can’t figure out why… There are no Natural error…
I am trying to set up tracing now. I set RPC trace level to 3 and also modified entirex.trace.properties to use level 3.
Is there anything else I have to do, because I still do not see the trace file?

Any help would be greatly appreciated.
Yelena

Hi Yelena,

For the RPC Server trace you should probably try starting the RPC server manually using the parameters

-TraceDestination

-TraceLevel [None|Standard|Advanced]

I think it is only possible to set these when you use the command line to start the server.

Regards,
Matthew

Matthew,

I am getting a log file now and all I can see is that when request is very short and simple - it does have a successful return like this:

M *** Convert to intern
*** Execute callnat… NATSERV2 at 12:12:54
*** Back from execute callnat NATSERV2 at 12:12:54
*** Convert to extern
*** Ready for reply to NATSERV2 at 12:12:54
********** USER-ID: SAG LIB : PROTO
*** Reply sent for NATSERV2 at 12:12:54

But when the request is more complex (calling to other subprograms and so on…) looks like by the time it’s complete (and I know that request is complete - I am saving the result) RPC is deregistering itself.

Log file in this case looks like this:
*** Convert to intern
*** Execute callnat… NATSERV2 at 12:13:04
*** Back from execute callnat NATSERV2 at 12:13:04
*** Convert to extern

“Convert to extern” is the last message and reply never send back.
.net client gets ‘00030010: EOC due to DEREGISTER of partner’ exception

Can it be a configuration problem? Any ideas???
I would really appreciate any help on this matter.

Yelena

what is your RPC TIMEOUT value set to in your server? Is the complex call taking longer than this value (the default is 55secs)?

You might also turn on the Broker trace to see what calls are going through the Broker and check if the Natural RPC Server is getting a Broker error that isn’t being shown for some reason.

We user entireX to communicate from the Web to the mainframe.
I got the same problem some time ago and on appon investigation found that I incorrectly declared a field as alpha on one place but as numeric on an other place in the module that communicates with EntireX.

So just go through your code an make sure the variables are the same every where.

good luck!