RE: Standby services on the primary - and shouldn't be
Date: Thu, 14 Mar 2013 23:42:37 +0000
Message-ID: <BD475CE0B3EE894DA0CAB36CE2F7DEB4010466A902_at_LITIGMBCRP02.Corp.Acxiom.net>
I kept thinking there's something stupid I'm missing and sure enough, there was. The act of forcing myself to write out the issue in more detail made me realize that registration was being done through the parameter REMOTE_LISTENER, which we always set to a TNS service name. I had validated that all 4 values were correct (2-node RACs, primary and standby) and also validated that tnsnames.ora had a proper entry for both the standby and primary (when checking today).
What I didn't realize is that through a bad combination of an incorrect entry in one of the tnsnames.ora files, inconsistent sync between the $ASM_HOME and $ORACLE_HOME copies, and a recent EM agent reinstall caused the standby database to pick up a value for REMOTE_LISTENER that pointed back to itself. I had found the bad tnsnames.ora copy 2 days ago, corrected it and made sure that all locations had the proper copy of tnsnames.ora, but that didn't change anything in what was registered with the listener on the primary. A few minutes back I altered REMOTE_LISTENER to be the same value (like "ALTER SYSTEM ... remote_listener='DHERRI_LISTENER'...") and that caused re-registration by the standby to it's own listener.
Thanks for being there to receive my emails and force me to think it through!!
Dave Herring
DBA
Acxiom Corporation
EML dave.herring_at_acxiom.com
TEL +1 630.944.4762
3333 Finley, Downers Grove, IL 60515, U.S.A
www.acxiom.com
-----Original Message-----
From: Herring Dave - dherri
Sent: Thursday, March 14, 2013 6:09 PM
To: Oracle-L (oracle-l_at_freelists.org)
Subject: Standby services on the primary - and shouldn't be
I've got a conundrum dealing with Oracle Net and data guard under 10g, where the standby service is registered with the primary listener and now the primary listener is trying to loadbalance between the 2, which obviously doesn't work under 10g.
Let's say the 2 databases are named "DHERRIA" (primary) and "DHERRIB" (standby), both being 2-node RACs. "DHERRIB" is a physical standby. When I check the listener on the primary I see the following:
LSNRCTL for Linux: Version 10.2.0.2.0 - Production on 14-MAR-2013 18:29:20
...
Services Summary...
Service "DHERRIA.us.dnb.com" has 2 instance(s).
Instance "dherri1", status READY, has 2 handler(s) for this service...
Instance "dherri2", status READY, has 1 handler(s) for this service...
Service "DHERRIA_DGB.us.dnb.com" has 2 instance(s).
Instance "dherri1", status READY, has 2 handler(s) for this service...
Instance "dherri2", status READY, has 1 handler(s) for this service...
Service "DHERRIA_DGMGRL.us.dnb.com" has 1 instance(s).
Instance "dherri1", status UNKNOWN, has 1 handler(s) for this service...
Service "DHERRIA_XPT.us.dnb.com" has 2 instance(s).
Instance "dherri1", status READY, has 2 handler(s) for this service...
Instance "dherri2", status READY, has 1 handler(s) for this service...
Service "DHERRIB_DGB.us.dnb.com" has 2 instance(s).
Instance "dherri1", status READY, has 1 handler(s) for this service...
Instance "dherri2", status READY, has 1 handler(s) for this service...
Service "DHERRIB_XPT.us.dnb.com" has 2 instance(s).
Instance "dherri1", status READY, has 1 handler(s) for this service...
Instance "dherri2", status READY, has 1 handler(s) for this service...
Service "DHERRIB.us.dnb.com" has 2 instance(s).
Instance "dherri1", status READY, has 1 handler(s) for this service...
Instance "dherri2", status READY, has 1 handler(s) for this service...
The 3 "DHERRIB*" sevices at the end are causing issues. Connections are regularly getting routed to the standby and then failing because it's in mount state. This even happens with RMAN operations on the primary where we're just cleaning up archivelogs but some channels get routed to the standby and then fail.
Is there any way to find out what's convincing the primary listener to pick up these standby services? Is there any way to stop them? I realize there's some communication between the primary/standby but in the other 4 DG environments I don't have this situation. On top of it this started recently and I can't find any changes that would cause this.
Dave Herring
DBA
Acxiom Corporation
EML dave.herring_at_acxiom.com
TEL +1 630.944.4762
3333 Finley, Downers Grove, IL 60515, U.S.A
www.acxiom.com
The information contained in this communication is confidential, is intended only for the use of the recipient named above, and may be legally privileged.
If the reader of this message is not the intended recipient, you are hereby notified that any dissemination, distribution or copying of this communication is strictly prohibited.
If you have received this communication in error, please resend this communication to the sender and delete the original message or any copy of it from your computer system.
Thank You.
--
http://www.freelists.org/webpage/oracle-l
Received on Fri Mar 15 2013 - 00:42:37 CET