Re: Oracle RAC listener on a 2nd network

From: Jeremy Schneider <jeremy.schneider_at_ardentperf.com>
Date: Wed, 30 Apr 2014 09:04:04 -0500
Message-ID: <CA+fnDAYGNZo0O0dfdj=TW7VG7CrTLDFeM1+VsEXf=65zTsPTOg_at_mail.gmail.com>



Just noticed this old thread.

FYI - I recently helped architect and manage an environment with a lot of databases where two networks (and multiple RAC dbs per cluster) were the standard for all deployments. There was a public network for client traffic but data guard traffic went over a backup network. I believe our internal processes were based on the note Raj mentioned [1063571.1]. This was all on 11.2.0.3 with recent PSUs applied and it works very well.

There's only one problem that I remember. When adding a new node to a cluster, the root.sh script which adds the node can't handle the extra network and fails. It creates the second network but doesn't add the VIP, then attempts to start the vips which fails. The workaround I found was to open a second terminal window and run the following command while root.sh is running:

srvctl add vip -n <NODENAME> -A <NODE-BVIP-IP>/255.255.254.0 -k 2

Other than that, the second network worked great for us on 11gR2. It's all managed by the GI clusterware so failover, listeners, etc work as they should.

-Jeremy

--
http://about.me/jeremy_schneider


On Wed, Apr 23, 2014 at 1:27 AM, Martin Bach
<development_at_the-playground.de>wrote:


> You are welcome.
>
> When I started putting my notes on WordPress I realised that the post's
> length got out of hand a bit. Next pieces to follow end of the week and
> next week. I had all of this written down in February but haven't come
> around posting it.
>
> Martin
> --
> Martin Bach
> Enkitec
> martincarstenbach.wordpress.com
-- http://www.freelists.org/webpage/oracle-l
Received on Wed Apr 30 2014 - 16:04:04 CEST

Original text of this message