Re: RAC and Dataguard
- From: Laimutis.Nedzinskas@xxxxxx
- To: oracle-l@xxxxxxxxxxxxx
- Date: Thu, 28 Mar 2013 08:29:23 +0200
it was a long time ago...
But here are excerpts from the scripts I use(d a long time ago):
To confirm your ideas about registering log files on standby:
PROMPT Redo gaps:
PROMPT
PROMPT if available copy and register all missing logs.
PROMPT to register use: ALTER DATABASE REGISTER PHYSICAL LOGFILE
'filespec1';
PROMPT
SELECT THREAD#, LOW_SEQUENCE#, HIGH_SEQUENCE# FROM V$ARCHIVE_GAP order by
THREAD#, LOW_SEQUENCE#;
PROMPT
PROMPT
PROMPT Highest log transfered:
PROMPT
PROMPT if available copy and register all higher logs per thread.
PROMPT to register use: ALTER DATABASE REGISTER PHYSICAL LOGFILE
'filespec1';
PROMPT
SELECT UNIQUE THREAD# AS THREAD, MAX(SEQUENCE#) OVER (PARTITION BY thread#)
AS LAST_SEQ from V$ARCHIVED_LOG;
PROMPT
PROMPT
The recover syntax
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE CANCEL;
or until sequence:
ALTER DATABASE RECOVER MANAGED STANDBY DATABASE NODELAY THROUGH THREAD 1
SEQUENCE &v_prim_seq;
And finally:
alter database recover managed standby database cancel;
or (?)
alter database recover database cancel;
ALTER DATABASE ACTIVATE STANDBY DATABASE;
or (interesting stuff and not applicable in your case because of SKIP
STANDBY LOGFILE)
REM !!!if the previous stuff fails then still trying this old good:
uncomment it!!!
REM ALTER DATABASE ACTIVATE STANDBY DATABASE SKIP STANDBY LOGFILE;
---------------------------------------------------------------------------------
Please consider the environment before printing this e-mail
From: Viljo Hakala <hakala.viljo@xxxxxxxxx>
To: Andrew Kerber <andrew.kerber@xxxxxxxxx>
Cc: "oracle-l@xxxxxxxxxxxxx" <oracle-l@xxxxxxxxxxxxx>
Date: 2013.03.27 22:19
Subject: Re: RAC and Dataguard
The redo logs were on a shared storage and visible to the second instance.
But I don't think the second instance would use the primary node's redo
logs and switch them over and deliver to the standby,
so the standby would have to be recovered from the unapplied, undelivered
redo logs if you don't want to lose data
in the primary node's redo logs.
And how to apply these undelivered redo logs was my question.
Andrew Kerber <andrew.kerber@xxxxxxxxx> kirjoitti 27.3.2013 kello 22.13:
> shared
--
//www.freelists.org/webpage/oracle-l
--
//www.freelists.org/webpage/oracle-l
Other related posts:
