Re: Oracle RAC newbie

Brian Peasland wrote:
> Mladen Gogala wrote:
> > Brian Peasland wrote:
> >>> It's usually a good practice to ask for references and check them.
> >>> Oracle
> >>> sales person will probably be able to compile the list of similar
> >>> configuration in your general area. If he can't - don't buy, you
> >>> definitely do not want to be one and only owner of that unique
> >>> configuration within a 100 miles radius.
> >>
> >> Why within a 100 miles radius? In today's Internet-connected world, I
> >> communicate with many DBA's all around the globe, even in this forum!
> >> I do not need to be physically colocated with the person I am
> >> exchanging knowledge with.

So _that's_ what those "urgent" newsgroup postings are all about! :-)

> >
> > Brian, support is the problem here. If you are the only one within a
> > wide area, the company that sold you the disk array will likely not have
> > a support person close by, so you will have to contact Elbonia for tech
> > support. Also, Oracle will not have anybody particularly well versed in
> > your neighborhood, so in case you need it, you will have to wait for
> > somebody to be flown into your area. This world is a global village, but
> > the local user groups are still there for a reason. That reason is
> > usually not marketing perpetrated by the Oracle Corp. and its affiliates
> > during annual events. Those local luser groups are organized for mutual
> > support purposes. Oracle usually tries to use luser attitude
> > readjustment tools on them.
> >
>
> I see your point...to a point. :)
>
> Even where I'm at, which is often referred to as geographically remote
> for those on either of the US coasts, we have support professionals from
> various vendors within a relatively short driving distance. But I've
> worked with vendors where the nearest hardware support technician was
> over 400 miles away, and we did have to wait for someone to be flown in.
> But I remember one time waiting for a techie from IBM to drive 75 miles
> and waiting for a techie from another vendor to fly in to fix some
> hardware. The one who flew in got there faster, mostly because the IBM
> techie decided we weren't high on his priority list that day. (As an
> aside, that issue was fun because we got IBM to admit they screwed up
> and violated the SLA of their support agreement).

The ironic corollary to this is, those managers who think they can get more
productivity by yelling get prioritization results.

>
> And it has also been my experience that on many occasions, all the
> techie ends up doing is calling someone, somewhere else, because he
> can't fix the problem! Local support is not always a panacea for what
> ails you.

Not Oracle, but there was a problem at LAX with a landing computer which cascaded flight delays/cancellations all over the country. The tech was at Torrance airport, just a few miles away - by severely overcrowed LA freeways. When she finally got there, the tech just rebooted... some people were real upset they couldn't figure out the root cause. Rumour has it trucks working on one of the runways messed things up. Verisimilitude there (my commute was once from next to the Torrance airport to next to LAX, so I can relate - things are much worse now).

>
> As for the local user's group...I am a big supporter of the local user's
> group. But I recently worked in a city where I could count the number of
> such Oracle professionals on one hand, most of which were new to Oracle.
> So networking on a local level was not beneficial to me. Having local
> people and local vendors to contact is not always going to fix one's
> problems. Having like configurations does not mean having like problems.

And even some big city user groups devolve into useless Agile programming advertisments.

> And if I had to wait for someone locally (within 100 miles) to
> implement the same technological solution as I am looking to put
> together, my business may suffer. Why should I wait for a company on the
> other side of the city to implement RAC? If RAC (or insert any other
> technology) is the correct solution to my meet my business requirements,
> I'm going to implement that technology whether I have local support or
> local configurations that match that.

Hmmmmm... well, if bleeding edge is correct, more power to ya. I have yet to be
convinced the "Why you probably don't need RAC" paper is obsolete. But then again,
I don't have your data or requirements :-)

More ominously, I have seen way too many sites left turning in the wind with technological dead-ends that seemed like good ideas at the time.

>
> And like I said, many of you would refer my geographic location as
> remote (the Dakotas). So for me, it has become a way of life to network
> with people from all over the US, and sometimes all over the world. The
> state I live in has less than 800,000 people living in it and the house
> I am building is less than 2 miles from a major Microsoft facility. How
> many Oracle professionals are going to be around my area? Very few. And
> how many in this area have RAC up and running? None. If my business
> needs dictate RAC as the correct solution, am I going to put the brakes
> on that project just because I do not have anyone in the area doing the
> same thing? A definite no to that one. However, when putting together
> the project and performing a risk analysis, this is one of the items
> that will be there. But I can tell you that it will not rate very high
> in the Impact category.

The thing is, so many don't perform the proper risk analysis. I haven't seen one in years. No wait - decades. The big gummint ones I've seen were obsolete before they were finished. Much better to just copy what big commercial companies did for similar environments, then you can pick successful implementations out from the unsuccessful.

jg

--
@home.com is bogus.
"Life is half delicious yogurt, half crap, and your job is to keep the
plastic spoon in the yogurt." - Scott Adams