Re: Patch levels for Exadata primary and non-Exadata physical standby databases

From: Jeremy Schneider <jeremy.schneider_at_ardentperf.com>
Date: Fri, 23 Oct 2015 09:50:55 -0400
Message-ID: <CA+fnDAbLAku+-oaY-5XZq_RoCKs+RH_jRZKOTpsZiN7rdi=v=A_at_mail.gmail.com>



Hi Laimsi,

My understanding is that the exadata "bundle patches" can actually be installed on any system. I think it's just another notch up from PSU's as far as including a few more known issues - inclusive of some exadata-specific stuff and some generic stuff. Actually, I bet that if we read through the list of bugs that are included in the latest exa-BP, we'll find that lots that aren't exadata-specific. I don't have a good documentation link or MOS link for this, but my understanding is that the exadata bundle patch doesn't trigger any special relinks or object replacements that would break any database running on a normal (non-engineered-system) server. And just like normal, code paths which are specific for exadata are simply ignored on non-exadata databases but nothing breaks.

Maybe someone else here knows of a MOS note where this documented. Or for a little extra assurance, you could always open an SR and just ask "is it safe to install exadata bundle patches on any non-exadata system."

I do know of one company where installing exadata-specific BP's is a standard practice for all non-exadata systems, in addition to the exadata ones. Even shops that don't have any exadata at all might consider switching from the PSU track to the exa-BP track if they want more aggressive patching of known issues (with both the associated risks and benefits of this). Personally I never started installing the exadata BP's until I started actually administering exadata racks - but now that I have learned about the exa-BP's, I'd give some consideration to the idea of switching from PSUs to the exa-BP's across the board anywhere - even if the shop didn't have any exadata systems at all.

-Jeremy

--
http://about.me/jeremy_schneider


On Fri, Oct 23, 2015 at 1:58 AM, Laimutis Nedzinskas
<laimutis.nedzinskas_at_statoilfuelretail.com> wrote:

> Hi
>
>
>
> What is the practice of patch levels for Exadata primary and non-Exadata
> physical standby databases?
>
>
>
> Metalink‘s 1265700.1 states that: “The net effect of this requirement is
> that non-Exadata systems must ultimately run Database Patch for Exadata,
> which is supported when the non-Exadata system is being used in a Data Guard
> environment with an Exadata system.”
>
>
>
> I personally read it: install Exadata patch onto a non-Exadata system.
>
> Is this so ?
>
>
>
>
>
>
>
> Thank you in advance,
>
> Laimsi N
>
>
>
>
-- http://www.freelists.org/webpage/oracle-l
Received on Fri Oct 23 2015 - 15:50:55 CEST

Original text of this message