RE: Questions re LOCAL_LISTENER & REMOTE_LISTENER

Rajeev,

No, it is due to misconfiguration and I am still struggling trying to understand things.

Bill Wagman
Univ. of California at Davis
IET Campus Data Center
wjwagman_at_ucdavis.edu
(530) 754-6208
-----Original Message-----
From: Rajeev Prabhakar [mailto:rprabha01_at_gmail.com] Sent: Friday, November 16, 2007 2:16 AM
To: William Wagman
Cc: oracle-l_at_freelists.org
Subject: Re: Questions re LOCAL_LISTENER & REMOTE_LISTENER

Bill

Is/Was there any sessions burst corresponding to the times when user connect to your instances ? If yes, it could be inability of linux kernel to
quickly adjust the load levels and in turn skewing the session distribution.

Please look into the following listener.ora parameter. Hopefully, it would
help.

PREFER_LEAST_LOADED_NODE_[listener_name] to ON or OFF

-Rajeev

On Nov 15, 2007 7:52 PM, William Wagman <wjwagman_at_ucdavis.edu> wrote:
> Greetings,
>
> I am still trying to get my head around this issue of LOCAL_LISTENER &
> REMOTE_LISTENER parameters. I am running a two node RAC cluster on
> RHEL4, Oracle 10.2.0.3.0 EE. ASM and DBMS are installed in separate
> oracle homes and the listener is running out of the $ASM_HOME. When
> pointing at the DBMS instance on each node show parameter listener
> returns the following...
>
> SQL> show parameter listener
> NAME TYPE VALUE
> ------------------------------------ -----------
> ------------------------------
> local_listener string (ADDRESS = (PROTOCOL
=
> TCP)(HO
> ST =
> brownshoe-vip.ucdavis.edu
> )(PORT = 1521 ))
> remote_listener string LISTENERS_MOTHRA
>
> On each node local_listener points to brownshoe-vip. The name of the
> second node is tenspeed (remember, old detectives?) and so I think on
> that node it should *instead* look like this...
>
> SQL> show parameter listener
> NAME TYPE VALUE
> ------------------------------------ -----------
> ------------------------------
> local_listener string (ADDRESS = (PROTOCOL
=
> TCP)(HO
> ST =
> tenspeed-vip.ucdavis.edu
> )(PORT = 1521 ))
> remote_listener string LISTENERS_MOTHRA
>
> The tnsnames.ora on each machine looks like this...
>
> # tnsnames.ora Network Configuration File:
> /opt/pkg/oracle/product/10.2.0/dbms/network/admin/tnsnames.ora
> # Generated by Oracle configuration tools.
>
> LISTENERS_MOTHRA =
> (ADDRESS_LIST =
> (ADDRESS = (PROTOCOL = TCP)(HOST = brownshoe-vip.ucdavis.edu)(PORT
=
> 1521))
> (ADDRESS = (PROTOCOL = TCP)(HOST = tenspeed-vip.ucdavis.edu)(PORT =
> 1521))
> )
>
> MOTHRA =
> (DESCRIPTION =
> (ADDRESS = (PROTOCOL = TCP)(HOST = brownshoe-vip.ucdavis.edu)(PORT
=
> 1521))
> (ADDRESS = (PROTOCOL = TCP)(HOST = tenspeed-vip.ucdavis.edu)(PORT =
> 1521))
> (LOAD_BALANCE = yes)
> (CONNECT_DATA =
> (SERVER = SHARED)
> (SERVICE_NAME = mothra.ucdavis.edu)
> )
> )
>
> MOTHR2 =
> (DESCRIPTION =
> (ADDRESS = (PROTOCOL = TCP)(HOST = tenspeed-vip.ucdavis.edu)(PORT =
> 1521))
> (CONNECT_DATA =
> (SERVER = SHARED)
> (SERVICE_NAME = mothra.ucdavis.edu)
> (INSTANCE_NAME = mothr2)
> )
> )
>
> MOTHR1 =
> (DESCRIPTION =
> (ADDRESS = (PROTOCOL = TCP)(HOST = brownshoe-vip.ucdavis.edu)(PORT
=
> 1521))
> (CONNECT_DATA =
> (SERVER = SHARED)
> (SERVICE_NAME = mothra.ucdavis.edu)
> (INSTANCE_NAME = mothr1)
> )
> )
>
> The issue I am seeing is that the majority of connections are going to
> node 1 (brownshoe) and I think this has something to do with it but
I'm
> still puzzled. Oracle is also telling me that I should change the
> tnsnames.ora so it looks like this on node 1...
>
> LOCAL_LISTENER_BROWNSHOE =
> (ADDRESS_LIST =
> (ADDRESS = (PROTOCOL = TCP)(HOST = brownshoe-vip.ucdavis.edu)(PORT
=
> 1521))
> )
> REMOTE_LISTENER_BROWNSHOE =
> (ADDRESS_LIST =
> (ADDRESS = (PROTOCOL = TCP)(HOST = tenspeed-vip.ucdavis.edu)(PORT =
> 1521))
> )
>
> And the mirror image on node 2. This is puzzling to me as the
> tnsnames.ora listed in example 9.1 of the Real Application Clusters
> Installation guide looks like the one I currently have. I think my
lack
> of understanding has to do with how the local and remote listener
> parameters function so if someone could help me in my understanding I
> would be most grateful.
>
> Thanks.
>
>
>
> Bill Wagman
> Univ. of California at Davis
> IET Campus Data Center
> wjwagman_at_ucdavis.edu
> (530) 754-6208
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>

--
http://www.freelists.org/webpage/oracle-l