Re: Upgrade issues 11.2.0.3 to 11.2.0.4?

From: MARK BRINSMEAD <mark.brinsmead_at_gmail.com>
Date: Tue, 31 Mar 2015 22:31:18 -0400
Message-ID: <CAAaXtLDNLuEjSorDSEVZ0O=L-9sBThGMFp_7gvZ4DxJ8EWcnRg_at_mail.gmail.com>



Absolutely!

There is no simple metric for how many databases a single DBA can manage. Although, if we phrase the question THAT way, I would say the number of databases a *single* DBA can manage is usually *zero*. You need a *team* of DBAs -- usually 3 or 4 at least -- in order to provide adequate coverage for:

  • sick time
  • vacation
  • on-call rotations
  • after-hours (planned) work

Forcing a DBA to work completely alone is the height of cruelty, I think.

But aside from that, there is really no generic way to say how much effort is involved in managing a database. You can have teeny-tiny QA databases that are hardly ever used, and probably require *no* effort to manage, and you can have great behemoth production databases that are so enormous and so complex (and supporting applications that are so poorly designed and written) that a "single" database -- with its standbys, clones, and staging areas -- might require a team of 30 DBAs and system administrators working around the clock to keep it operating acceptably.

In any case, there is a huge variation in the tasks, effort, and quality of service, needed to "manage a database" as you move from one organization to another, from one environment to another, and even from one application to another. This is the reason why it is (or at least, *should be*) so hard to get anybody to commit to "fixed-price" contracts for database management. And it is the reason why there is no single correct answer to the question.


The question posed in the "other forum" came from an HR person, and it was really rather silly, and quite easy to answer, at least superficially.

The question was:

*Is it possible for a DBA to manage more than one database?*

Worded as it was, the answer is simple:

*Of course it is possible. Don't be silly. But it might depend on the database.*

The question becomes more interesting, though, if you consider that the original poster may not have understood the question she was asking. Perhaps the question was meant to be:

*Is it possible for a DBA to manage more than one database product?* (i.e., Oracle, SQLserver, MySQL, DB2 ...)

Technically, you could still give the same answer. It's silly to ask whether something is *possible*. But what if we change that to *"practical*" ?

Now there is potential for some interesting discussion. Is it better to have a single team of DBAs who manage all databases across all products and platforms within an organization (or within a business unit)? Or is it better to SPLIT that team, and have multiple smaller teams each specializing in a single technology?

I already know my answer. And those who can find the "other forum" will be able to see it -- although probably not stated nearly as well as I would here. (The "other forum" is simply not worth the time and effort.) But answering a question like this in *this* forum is likely to spark a flame-war of scriptural proportions -- so I will bow out without answering, and watch for the war to begin!

Who knows? If the question takes we might see an interesting discourse. Or not.

Cheers!

On Tue, Mar 31, 2015 at 9:04 PM, Mladen Gogala <dmarc-noreply_at_freelists.org> wrote:

> On 3/31/2015 6:06 PM, MARK BRINSMEAD wrote:
>
>> I have been engaged in a discussion on another forum concerning how many
>> databases a small group of DBAs can manage. Some say the answer is "one",
>> while I say the answer is -- or can be -- "hundreds". But 5 DBAs are not
>> going to manage 200 or 300 databases if they all need to be patched once
>> every 90 days, and we need to do 2 weeks of testing for each one prior to
>> patching. (And some would say that "only" two weeks testing is wholly
>> inadequate.)
>>
> Not all databases are created equal, some are more equal than the others.
> My sandbox DB will, of course, be regularly maintained and patched up to
> the hilt, while the critical business database may not be so lucky. The
> reason is that patching up the latter requires a bit more testing, things
> like change control and management approval, while my sandbox can be shut
> down the very moment that I finish downloading the latest PSU, one which is
> coming up in April. There are usually 5 or 6 databases having critical
> status and upgrading them is always a project. Upgrading development
> databases, QA, UAT or DBA sandbox database is much easier.
> Upgrading critical production databases is always a project and surprises
> can end up as resume generating events.
>
>
> --
> Mladen Gogala
> Oracle DBA
> http://mgogala.freehostia.com
>
> --
> http://www.freelists.org/webpage/oracle-l
>
>
>

--
http://www.freelists.org/webpage/oracle-l
Received on Wed Apr 01 2015 - 04:31:18 CEST

Original text of this message