Oracle announced Oracle DB Appliance today. This is a deja vu, from the
era of the Oracle Network Computer when Oracle was the first to offer a
"database appliance". NC and its brethren database appliance died around
the year 2000, which means that the early bird doesn't necessarily catch
the worm.
This time, however, things are different. There are several similar
machines around, even if we don't count Exadata and Exalogic boxes. There
is Teradata Kickfire, there is IBM's "virtual DB2 appliance", a very
innovative piece of software that I still count as an appliance and, of
course, Netezza DW appliance.
Apparently, the trend is to sell the database, along with the hardware
needed to run it and also have the company that sold it to administer it
as a part of the support.
IBM has the most experience with it, since they're selling DB2 on the
mainframe systems for a very long time. The whole mainframe system can be
regarded as one huge DB2 appliance.
That is the usual characteristic of the db appliance market: appliances
are high end contraptions. Exadata, Netezza, Kickfire or mainframes are
very expensive and glamorous contraptions that with at least six figure
price tags and are unlikely to be encountered in small to medium size
companies. Furthermore, no matter how fast these things are, the state of
technology is simply such that these things will always require a good
DBA. Oracle is well positioned to win the inevitable DB appliance wars,
but will not impact the vast majority of the Oracle users.




--
http://mgogala.byethost5.com

I respectfully disagree, at least in the DW arena...

Cheers.

Carlos.

On Thu, 22 Sep 2011 08:46:41 -0700, CarlosAL wrote:

Any particular reason for disagreement? Oracle is the largest DB vendor
and as such is much, much bigger than Teradata, Netezza, Greenplum and
Vertica combined. Contrary to some platitudes, the size does matter, at
least when it comes to the DB companies.



--
http://mgogala.byethost5.com

Mladen.

This can be true from a 'manager point of view'. When it comes to pure
technical features (and even the performance/cost) we are talking
another language: Toyota may be the largest car-selling company, but
I'd rather buy a... (put here the car you like most).

Anyway, I admit I'm biased here (just like I was -the other way- some
years ago) and I don't think this is the place for a very personal
discussion.

Cheers.

Carlos.

On Sep 23, 1:36*am, CarlosAL <miotromailcar... (AT) netscape (DOT) net> wrote:
I agree, Exadata v. all those others would wrest some of the market
from them, but between general dislike of Oracle and specific DW
requirements, not fatally. But the appliance announced? The only way
I imagine is perhaps some analogue of the way Google or Amazon
leverage quantities of commodity servers, and really I would only see
Amazon doing something like that, anyone else would just use the
commodity servers, anything large would have to have RAC expertise
available anyways. One would think.

Of course, once you start talking about specific DW requirements, then
you bring the whole hawhoopdedoop Stonebraker crowd, and there small
org size is probably an advantage, being easily able to adapt to the
buzz of the moment.

Aw, this is usenet, spill your most personal secrets to crowds of
anonymous strangers ;-)

jg
--
@home.com is bogus.
http://www.signonsandiego.com/news/2...dnt-share-inf/

On Fri, 23 Sep 2011 09:29:18 -0700, joel garry wrote:

I don't see the appliance as very successful on the low end. It's still a
hefty two node RAC, licensed per CPU. Exadata is rather successful
because it's a racing car: a high end contraption built for speed, with
the premium price. Appliance looks like a crossbreed: body of a Ferrari
and engine of Ford Taurus. It's neither as fast as a Ferrari nor it has
the luggage space of a Taurus. This is not the first "Oracle Database
Appliance". Hopefully, this time it'll fare a bit better.



--
http://mgogala.byethost5.com

On Sep 23, 12:26*pm, Mladen Gogala <gogala.mla... (AT) gmail (DOT) com> wrote:
I think we are agreed, although I was wondering if perhaps there is
some market of linux/rac clusters looking for an upgrade less than
hexadata. Thinking on that, there might be a lot of those, people who
drank the kool-aid 4-5 years ago. Those who can mine support contract
info might have seen that and gone hmmm, turn those sour old 30-hour-
configuring-and-still-not-working lemon-memories into let's-kick-some-
life-into-the-old-Sun-we-got-on-the-cheap-ade. Just add marketing
sugar and stir the blogosphere.

jg
--
@home.com is bogus.
Oh, I thought they said RAT In The Box http://youtu.be/dY-FOI-9SOE

Mladen:

# I don't see the appliance as very successful on the low end. It's
still a hefty two node RAC, licensed per CPU.

Well the announcement was rather vague but strong hints that you don't
need to run RAC on it and you can migrate existing licenses over to it
and you can pay for as much of it as you want ( or are using up ).

Price tag for a server where you are not using all of it seems pretty
high.

Remember they kept ( or at least Mark Hurd did ) pushing the words
SMB. A partner based solution for the partners to sell this into
small businesses and toss it over the fence without having to work
hard to configure it and/or support it ... seems like.

Mladen Gogala wrote,on my timestamp of 24/09/2011 5:26 AM:

It's not expandable, it cannot be upgraded (patching software is NOT
upgrading!), it cannot scale horizontally and it doesn't have enough capacity to
be a single storage solution for even a mid-sized business.
Looks to me like a fail..
Yeah, sure: lots of blogging about it. What else would be expected? It's a
marketing campaign, not a product.

Nuno:

# it cannot scale horizontally

Neither can my dedicated database servers ... well I guess they can if
we buy a different one!

# and it doesn't have enough capacity to be a single storage solution
for even a mid-sized business.

??? It has a lot more than we use ... guess you may have a pretty big
idea of what a mid sized business is eh?

Is SMB different from your mid sized business concept?

# Looks to me like a fail.. Yeah, sure: lots of blogging about it.
*What else would be expected? It's a marketing campaign, not a
product.

Well you need to have a product for a marketing campaign right?

Sure seems like it is both a product and a marketing campaign. Or at
least it will be a product when they are officially on sale ...
probably next week?

I was expecting a bigger announcement though the way the web
announcement was pushed.

I was guessing that Oracle was announcing they were buying HP!