Re: Upgrading with no patches in the "base"?

What this says to me, is that vendor is trying to see if an unpatched home will work, and might lead to not patching at all, because it works, patching would mean an extra risk. If not, I cannot see any valid reason for trying/testing or otherwise running on a version you never intent to run on.

It might come from the old (approximately 15 years old that I’ve seen it) practice of installing software and configure it to run, and then leave it be. I don’t think it ever was the best way of doing it, but it was really common the aforementioned time ago.

In todays world, there is not a single reason to create a patching strategy that fits your company and stick to that. Security is really prominent for the past 10 years or so. But also for supportability: if you run into an issue or a bug, one of the first things support will ask is to patch it to current and see if the problem persists. You could argue if that is a good practice, but I think it’s sensible to apply patches to get issues solved and vulnerabilities fixed you might have never suffered, but as a result will not run into. Of course there is a downside too: things might work differently.

Frits Hoogland

http://fritshoogland.wordpress.com <http://fritshoogland.wordpress.com/> frits.hoogland_at_gmail.com <mailto:frits.hoogland_at_gmail.com> Mobile: +31 6 14180860

> On 8 Jan 2021, at 15:29, Michael Kline <mkline1_at_comcast.net> wrote:
>
> Hearing that an application is going to be upgraded from 12.1 to 12.2.
>
> Vendor is saying they will create a “blank, no patched” 12.2 $ORACLE_HOME, and then upgrade the database.
>
> They will test this for a while, and if everything is fine, THEN they will apply the patch.
>
> I’ve never heard of such a thing and have been working on Oracle databases since 1983, version 4.0.
>
> Is there logic in this? We try to keep all databases at N-1 on patching.
>
>
> Michael Kline

--
http://www.freelists.org/webpage/oracle-l