Re: lsnrctl doesn't really stop the listener

On Feb 5, 10:22 pm, "Ramon F Herrera" <r..._at_conexus.net> wrote:
> Oracle9i Enterprise Edition Release 9.2.0.1.0 on Linux
>
> I have been trying to stop the listener through the usual way:
>
> LSNRCTL> stop
> Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=IPC)(KEY=EXTPROC)))
> The command completed successfully
>
> But very often, the listener stays up. I can see it with 'netstat -a',
> I can telnet into port 1521 and I can run sqlplus from another
> machine. In those cases in which the listener refuses to die, I can
> see that there are two processes talking to each other through that
> port (with netstat -ap): tnslsnr and ora_pmon.
>
> I don't have my 10g available right now, so I can investigate this
> matter there.
>
> Is this zombie-like behavior normal? Can some of you folks check
> whether your listeners stay up after they are stopped?
>
> Should I just kill 'tnslsnr' by hand when lsnrctl fails?
>
> TIA,
>
> -Ramon F Herrera

9.2.0.6, Solaris 10:

/export/home/oracle> lsnrctl stop

LSNRCTL for Solaris: Version 9.2.0.6.0 - Production on 06-FEB-2007 14:53:49

Copyright (c) 1991, 2002, Oracle Corporation. All rights reserved.

Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=TCP)(HOST=********) (PORT=****)))
The command completed successfully
/export/home/oracle> netstat -ap
Net to Media Table: IPv4
Device IP Address Mask Flags Phys Addr

------ -------------------- --------------- ----- ---------------
dmfe0  *****                255.255.255.255       ############
dmfe0  ****                 255.255.255.255 SP    ############
dmfe0  *****               240.0.0.0       SM    ##########

/export/home/oracle> ps -ef | grep lsnr
/export/home/oracle>

This might be a very good reason to purchase a support agreement and patch to at least 9.2.0.6.

David Fitzjarrell Received on Tue Feb 06 2007 - 08:59:12 CST