Re: typical TAF failover times?

From: DA Morgan <damorgan_at_psoug.org>
Date: Thu, 07 Feb 2008 09:32:20 -0800
Message-ID: <1202405515.686104@bubbleator.drizzle.com>


mh_at_pixar.com wrote:
> First, just to make sure I understand: All TAF configuration is
> in the tnsnames.ora entry, and doesn't need any setup on the
> RAC servers. Is that correct?

Depends on the version.

> Second, what are some typical TAF failover times? I'm in a
> gig-e LAN environment, so there won't be any serious network
> timing fluctuations.

Depends on the operating system.

> For our application it would be preferrable to fail over
> as quickly as possible, since we are servicing nodes in
> a pretty large compute farm.

Can you think of anyone that desires a slow failover?

> Would retries=3,delay=5 be reasonable? Then work
> should pick up in 15 seconds or so after a node fails.
>
> Many TIA!
> Mark

Typical is delay 5 and retries 180 but that does not mean either is suitable for your environment.

Provide additional information and perhaps we can get you better answers. Also, why TAF rather than FCF?

-- 
Daniel A. Morgan
Oracle Ace Director & Instructor
University of Washington
damorgan_at_x.washington.edu (replace x with u to respond)
Puget Sound Oracle Users Group
www.psoug.org
Received on Thu Feb 07 2008 - 11:32:20 CST

Original text of this message