Re: [tao-users] Question with respect to CORBA callbacks

[Douglas C. Schmidt]

Hi,

Thanks for using the PRF!

> TAO VERSION: 1.6.1
> ACE VERSION: 5.6.1
>
> HOST MACHINE and OPERATING SYSTEM:
> If on Windows based OS's, which version of WINSOCK do you
> use?:
> Red Hat Enterprise Linux 5.1
>
> TARGET MACHINE and OPERATING SYSTEM, if different from HOST:
> COMPILER NAME AND VERSION (AND PATCHLEVEL):
> Red Hat Enterprise Linux 5.1 ( GCC 4.1.1 )
>
> THE $ACE_ROOT/ace/config.h FILE [if you use a link to a platform-
> specific file, simply state which one]:
> ace/config-linux.h
>
> THE $ACE_ROOT/include/makeinclude/platform_macros.GNU FILE [if you
> use a link to a platform-specific file, simply state which one
> (unless this isn't used in this case, e.g., with Microsoft Visual
> C++)]:
> platform_linux.GNU
>
> CONTENTS OF $ACE_ROOT/bin/MakeProjectCreator/config/default.features
> (used by MPC when you generate your own makefiles):
>
> AREA/CLASS/EXAMPLE AFFECTED:
> Usage Question. How to close sockets of deactivated servants.
>
> DOES THE PROBLEM AFFECT:
> COMPILATION?
> No
> LINKING?
> No
> EXECUTION?
> Yes, it doesn't appear that after deactivating my callback
> that sockets are returned to the system.

Right - TAO decouples management of sockets from management of servants.
You can probably get the behavior you need by playing around with TAO's
connection caching options. Please see

TAO_ROOT/docs/Options.html#TDRF

for more discussion about these options.

Thanks,

Doug

> OTHER (please specify)?
>
> SYNOPSIS:
> I have an application which connects to TAO corba servers
> and registers a callback. It does some monitoring, then
> connects to another remote server, and so on and so forth.
> The problem I'm having is that I am calling deactivate_object
> via the POA on the derived class, and CORBA::release()
> on the pointer returned from the ->_this() call, but the
> socket for the callback is not being returned to the system.
> I am using the RootPOA, and monitoring socket usage via
> ls /proc/<pid>/fd | wc -l
>
> DESCRIPTION:
> From my research it appears that the RootPOA is designed to have
> servants activated and deactivated. Once the servant is
> deactivated and the reference count is 0 should the socket be
> closed for that servant? Is there a way to trigger this
> behavior?
>
> This causes problems because the point of my app is to connect,
> and register a callback with many different CORBA servers. The
> app then queries some info and disconnects moving on to the next
> server. I appear to be continually eating sockets when I register
> my callback with a new server.
>
> Is there a way to *delete* a servant that will no longer be
> used ( instead of just deactivating it ) from the POA so
> that any send/receive sockets it has open will be closed?
>
> REPEAT BY:
> 1. Look at the /proc/<pid>/fd count of your application
> 2. Create a callback via the ->_this() call on a servant derived class
> 3. Register that callback with a server
> 4. <receive a callback>
> 5. Deactivate the derived class via POA->deactivate_object
> 6. CORBA::release() the pointer returned from the ->_this() call
> 7. The /proc/<pid>/fd count of your application has increased by 1
>
>
>
> SAMPLE FIX/WORKAROUND:
> [If available ]
>
>
> Thank you for your time,
> Charles
>
>
>
> _______________________________________________
> tao-users mailing list
> tao-...@list.isis.vanderbilt.edu
> http://list.isis.vanderbilt.edu/mailman/listinfo/tao-users