Need help diagnosing Local OS authentication issue when running OracleXE Linux

[Adrian Penisoara]

Hi,

I am working on a FreeBSD port for the OracleXE for Linux package
and I stumbled upon a collateral issue when testing the functionality
of OracleXE: I am unable to manually shutdown the TNS Listener
(running "lsnrctl stop"), since the listener's "Local OS
authentication" appears to malfunction.
This feature prevents any other OS user but the user who initially
started the listener to execute administrative tasks on the listener,
including shutting down the listener. The listener process is a
background service and accepts commands through the standard 1521 TCP
port from the administration tool. I guess there is a problem when
"emulating" the Linux kernel/library calls related to OS credentials
for network sockets (?).

Does anyone else have clues on this ?
What would be the best way to proceed analyzing the issue -- e.g.
how should I trace the execution for the running listener process (and
eventually compare to a trace ran on a real Linux machine) ?

Thank you,
Adrian Penisoara
Ady (@freebsd.ady.ro)

[Chagin Dmitry]

On Thu, Oct 23, 2008 at 01:38:34AM +0200, Adrian Penisoara wrote:
> Hi,
>
> I am working on a FreeBSD port for the OracleXE for Linux package
> and I stumbled upon a collateral issue when testing the functionality
> of OracleXE: I am unable to manually shutdown the TNS Listener
> (running "lsnrctl stop"), since the listener's "Local OS
> authentication" appears to malfunction.
> This feature prevents any other OS user but the user who initially
> started the listener to execute administrative tasks on the listener,
> including shutting down the listener. The listener process is a
> background service and accepts commands through the standard 1521 TCP
> port from the administration tool. I guess there is a problem when
> "emulating" the Linux kernel/library calls related to OS credentials
> for network sockets (?).
>

Hi,

yes, this problem well-known to us :) please, see kern/102956 for
full description. I think that this PR will be closed soon.

> Does anyone else have clues on this ?
> What would be the best way to proceed analyzing the issue -- e.g.
> how should I trace the execution for the running listener process (and
> eventually compare to a trace ran on a real Linux machine) ?
>

--
Have fun!
chd