Re: RAC dataguard - what happens when recovering instance goes down ?

From: Karl Arao <karlarao_at_gmail.com>
Date: Mon, 18 May 2009 21:20:29 +0800
Message-ID: <12ee65600905180620j663005e9x711c616103c5cc17_at_mail.gmail.com>



Hi Prem,

I think it will just terminate the recovery, it may also corrupt the archivelog that it is currently being transported to the standby (which happened twice on me)..

And it depends on your data protection mode..

  1. Maximum Performance - you could configure it to use ARCn or LGWR for redo transport with async network IO.. if the standby database goes down, your primary still runs and you may incur archive log gap
  2. Maximum Availability - you configure it to use LGWR for redo transport with synchronous network IO.. if the standby goes down processing will continue on the primary database.
  3. Maximum Protection - you configure it to use LGWR for redo transport with synchronous network IO and AFFIRM attribute.. if the standby goes down.. the primary goes down as well..

On Mon, May 18, 2009 at 8:58 PM, Prem Khanna J <premj_at_rediffmail.com> wrote:

> Friends,
>
> we have 2 node 10gR2 RAC dataguard setup on AIX5.3L.
>
> Just curiuous:
> what happens when the recovering standby instance goes down ?
> Has anyone come across this situation ? RTFM-ed , Google-ed ..
> but still could not find the answer !
>
> Regards,
> Prem
>
> <http://sigads.rediff.com/RealMedia/ads/click_nx.ads/www.rediffmail.com/signatureline.htm_at_Middle?>

--
http://www.freelists.org/webpage/oracle-l
Received on Mon May 18 2009 - 08:20:29 CDT

Original text of this message