RE: Recovering a standby database with rman
Date: Fri, 28 May 2021 12:53:28 -0400
Message-ID: <51d801d753e1$fc352f90$f49f8eb0$_at_rsiz.com>
Before either RMAN or standby were products, the way we did this was to pause recovery, backup on the standby, end recovery on the standby, then switch on the primary, transmit the archived redolog result of that switch, then back up the redologs we had on the standby, save the controlfile, and finally resume recovery on the standby (unless apply was intentionally lagged). With RMAN of course that dance with pause and resume recovery shouldn’t be needed.
Are you doing a “physical backup” or using RMAN?
Without checking I would not presume that is how 19 and recovery on the standby works in their product 30 years later. I’m only answering because possibly everyone has flown the coop for the holiday weekend and I saw this. Having not actually done a commercial scale back up and test recovery in a couple decades, I’m mostly just confirming that it does seem like you’d need that piece. I suspect that RMAN will tell you that you need it, and that physical recovery will not complete normally without it.
I am mystified that this wouldn’t be the first example case study in the online documentation.
Good luck, I bet someone on the list (still) does this or something very similar every day.
mwf
From: oracle-l-bounce_at_freelists.org [mailto:oracle-l-bounce_at_freelists.org] On Behalf Of Howard Latham
Sent: Friday, May 28, 2021 5:25 AM
To: ORACLE-L
Subject: Recovering a standby database with rman
-- http://www.freelists.org/webpage/oracle-lReceived on Fri May 28 2021 - 18:53:28 CEST