Comparison of DB2 and Oracle?

[Rhino]

One of my friends, Scott, is a consultant who doesn't currently have
newsgroup access so I am asking these questions for him. I'll be telling him
how to monitor the answers via Google Newsgroup searches.

Scott has heard a lot of hype about DB2 and Oracle and is trying to
understand the pros and cons of each product. I'm quite familiar with DB2
but have never used Oracle so I can't make any meaningful comparisons for
him. He does not have a lot of database background but sometimes has to
choose or recommend a database to his clients.

Scott has enough life-experience to take the marketing information produced
by IBM and Oracle with a grain of salt and would like to hear from real
DBAs, especially ones who are fluent with both products, for their views on
two questions:

1. What are the pros and cons of the current releases of DB2 and Oracle?

2. What other sources of *independent* information are available to help
someone new to databases choose between DB2 and Oracle?

This is *not* a troll and we don't want to start a flame war! Scott just
want some honest facts to help him decide which product is best at which
jobs.

--
Rhino

[DA Morgan]

michael newport wrote:

>>Linux, Apache and PHP are succesful because there is a strong developer
and user community. Ingres doesn't have this, and making something
OpenSource doesn't cause this community to automatically build.

Linux, Apache and PHP did not start off successful. They grew.

Ingres has existed for a long time, the base IS there
comp.databases.ingres

The "base" is database developers not people that write kernel code in
C. They will all die of old age before they figure out how to give the
Ingres kernel capabilities that were in Oracle 8i.

>>Linux in particular benefited from the focus companies like Oracle, IBM
and others placed on it. The same level of focus is unlikely to happen
for Ingres.

Companies focus on Linux because it is free. A huge advantage.

Nonsense. Absolute ignorant nonsense. I consult for a division of The
Boeing company. The cost of an operating system compared to the total
cost of an application is so small as to be invisible. Do you really
think we are going to build a $15,000,000 system and worry about the
lousy few hundred or few thousand dollars for the O/S?

We chose Linux because it gave us better performance, in lab tests with
our application than did Win2K, WinXP, Solaris 2.9 and HP/UX 11i.

You really are out of touch with reality.

Oh and by the way Ingres is also free.

Oh and by the way ... we really don't give a damn. And neither
does anyone else.
--
Daniel A. Morgan
University of Washington
da******@x.washington.edu
(replace 'x' with 'u' to respond)

[DA Morgan]

Serge Rielau wrote:

Michael,

This thread is now very much off topic.
Comparison between Oracle and Ingres should be in these respective
newsgroups if you insist on having these debates.
I don't see where comp.databases.ibm-db2 is relevant here.

Cheers
Serge

I agree. I am kill-filing this thread. Newport is just trolling and I
am tired of responding to a neolithic fanatic. You'll so no more on
this from me.
--
Daniel A. Morgan
University of Washington
da******@x.washington.edu
(replace 'x' with 'u' to respond)

[michael newport]

Not once in this entire thread have Oracle Report Server (it doesn't

even exist any more) or JAVA been part of any post. Don't try to change
the subject. That is a activity best left to small children.

Daniel,

in Ingres I wrote 4GL, in Oracle I write PL/SQL

Because you don't know how to write Java in the database?

in Ingres I wrote SQL, in Oracle I write SQL

Because you don't know how to write Java in the database?
Daniel,
it seems that you have short term memory loss,

it also seems that you have never used Ingres.

Regards
Michael Newport

[michael newport]

Serge,

I already did an oracle vs ingres thread but was disappointed with the
answers that I received.

Please notice that I only commented on databases that I have
experience of,
unlike our favourite troll Mr.Morgan who has a problem with anything
not spelt Oracle.

Regards
Michael Newport

[Darin McBride]

DA Morgan wrote:

>You can, if you wish, get support for a license. That is annual, and
provides unlimited support calls. Quite different from licensing.

<heavy sigh>

When the facts don't support a position it is not uncommon to have
pseudofacts invented.

To be really fair, this is just nitpicking on the semantics. To most
people, "licensing" really means "total cost payable to the vendor."
If I must buy an annual support contract to ensure the product's
success in my environment, that's the same thing, at the end of the
day - money going from my company to Oracle, IBM, MS, CA, whatever.

Most of the thread arguing against Ingres has talked vaguely about a
"total cost" of ownership - and the one time Michael has a valid point
about the vendor portion of that TCO (which, of course, does not
validate anything else he says), you go and nitpick his terminology.

Then again, perhaps it's not uncommon that when your opponent is
generally making no sense, that you stop reading his posts objectively,
and just assume that the whole argument is absurd, rather than just the
individual (and overwhelming) portions of it that really are absurd?

[HansF]

Darin McBride wrote:

>
To be really fair, this is just nitpicking on the semantics. To most
people, "licensing" really means "total cost payable to the vendor."
If I must buy an annual support contract to ensure the product's
success in my environment, that's the same thing, at the end of the
day - money going from my company to Oracle, IBM, MS, CA, whatever.

You seem to separate vendor cost from internal cost.

The implication is that any development costs or maintenance costs resulting
from inventing software to compensate for capabilities not in Ingres are
not to be counted, but paying Oracle or IBM for those same capabilities are
to be counted. I'd be concerned about that style of accounting - it's
quite reminiscient of the CapEx vs OpEx accounting invented to get around
regulations in some industries.

No matter which way we try to wiggle, companies need to manage total cost,
not just "total cost payable to the vendor." The proof is in the attempts
at outsourcing - whether it works well (or works at all) is irrelevant, the
relevance is that companies are doing this (in desparation?) to get total
costs under control.

However, as Serge says, this thread is WAY off topic and no longer relevant.
If you want to continue this discussion, I suggest we go to some Ingres or
open source advocacy group. I hereby stop responding to the Ingres and
Open Source discussion in this thread and apologize to all for not having
stopped sooner.

/Hans

[michael newport]

Then again, perhaps it's not uncommon that when your opponent is

generally making no sense, that you stop reading his posts objectively,
and just assume that the whole argument is absurd, rather than just the
individual (and overwhelming) portions of it that really are absurd?

which bit did you have trouble with ?

[michael newport]

Jean-David Beyer <jd*****@exit109.comwrote in message news:<10*************@corp.supernews.com>...

michael newport wrote:

>Well, you need to get more experience with new stuff. Doing the same
thing over in a different environment should give you an increased
appreciation of what you are doing, and what you could be doing.

It did, and the similarities were all too obvious.

>>>That's not the fault of the product. That direct and proximate
responsibility falls on you for being a dinosaur. How much code have
you implemented with bulk binding? How much with the model clause?
How much with analytic functions? How many materialized views with
refresh logs?

its answers the users needs.
and it was written by the dealine.
which meant my company got paid.
although some of this money was then sent to Oracle to pay for the
licence.
if we had used Ingres we could have done the same job for less, or
increased our profits.

I used to work for a vendor of a product that worked on multiple
databases, including Ingres. They dropped Ingres support due to lack
of interest from potential customers. Are you sure whoever paid your
company would have been interested with Ingres? Many products are
considered more desireable simply because they are more expensive.
Stupid, true, but the way of the world.

I agree that CA sales and marketing were bad. But Ingres the product is not.
CA also wasted time and money on speculative products like Jasmine and Opal.
Linux / Apache / PHP have taken off because they are reliable and OpenSource.
I predict the same for Ingres.

I would be curious what the advantages of Ingres might be over other free
(depending on exact usage) dbms's such as postgreSQL and MySQL. I know
that Ingres has been around since even before Oracle existed (late
1970s?). I suppose postgreSQL is a descendant of Ingres.

For desktop use, it probably matters little, though after fussing around
with a bunch of them, I chose to pay IBM for their DB2 UDB because it just
plain worked better and they seemed to follow standards (such as for
Embedded SQL) better than did Informix or postgreSQL did at the time I
tried them (mid to late 1990s).

Open Source is good, and not just Ingres.
But I used Ingres for a long time, and I know it works.
Oracle also works but costs a lot of money.

I also read that IBM and Sybase appear to be going opensource.

[Serge Rielau]

michael newport wrote:

I also read that IBM and Sybase appear to be going opensource.

IBM is a company, not a product.
IBM Cloudscape has been open sourced as "Derby".
There aren't even rumours that IBM may open source one of it's
mainstream commercial RDBMS (DB2, IDS, XPS and RedBrick)

Cheers
Serge

[michael newport]

Serge,

would you like to see these other IBM products OpenSourced ?

Regards
Michael Newport

[michael newport]

>michael newport wrote:

>

Daniel,

what do you do at the University of Washington ?

nothing to do with education ?

Regards
Michael Newport

Teach databases something that might have interested you
once in your life.

I am still interested, which is why we are having this discussion.
But rather than back a product because it has a particular brand,
I prefer a more realistic discussion of experience.

Have you ever used Ingres ?

I don't "back" a product. I work routinely in multiple products. That
I teach one relates to what the university's students want ... not what
I do.

But no one wants to learn Ingres. It is a decaying corpse that CA has
attempted to bury at sea. If you want to work with a real open-source
database the clear choice is MySQL.

So you have never used Ingres.

I guess the OpenSource tide brought Ingres back to shore.

Why is the clear choice MySQL ?

[Noons]

"Jim Kennedy" <ke****************************@attbi.netwrote in message news:<Tzifd.6802$HA.6215@attbi_s01>...

If this is the same Ingres I used awhile ago I wouldn't touch it with a ten
foot pole even if you paid me. The concurrency model sucks, start a
transaction, insert a record, lock 95% of the table if it has a primary
key - because the page locks on the index locks most of the pages. NO ONE
ELSE COULD GET ANY WORK DONE, unless you threw out the transaction model and
went to auto commit. POS.

The very same POS. That got shafted out of the market
PRECISELY because of the crap it always was.
Did you ever try crashing the server? Best way to ensure
you lost all your work, with Ingres.

[Serge Rielau]

You are asking me whether I want to be fired from my current job.
I currenly own a good portion of DB2 for LUW's SQL Compiler code.
For sure I'm not in it for the charitable work although it sometimes
feels like it.

Cheers
Serge

[michael newport]

DA Morgan <da******@x.washington.eduwrote in message news:<1098928130.887686@yasure>...

michael newport wrote:

>Linux, Apache and PHP are succesful because there is a strong developer
and user community. Ingres doesn't have this, and making something
OpenSource doesn't cause this community to automatically build.

Linux, Apache and PHP did not start off successful. They grew.

Ingres has existed for a long time, the base IS there
comp.databases.ingres

The "base" is database developers not people that write kernel code in
C. They will all die of old age before they figure out how to give the
Ingres kernel capabilities that were in Oracle 8i.

>Linux in particular benefited from the focus companies like Oracle, IBM
and others placed on it. The same level of focus is unlikely to happen
for Ingres.

Companies focus on Linux because it is free. A huge advantage.

Nonsense. Absolute ignorant nonsense. I consult for a division of The
Boeing company. The cost of an operating system compared to the total
cost of an application is so small as to be invisible. Do you really
think we are going to build a $15,000,000 system and worry about the
lousy few hundred or few thousand dollars for the O/S?

We chose Linux because it gave us better performance, in lab tests with
our application than did Win2K, WinXP, Solaris 2.9 and HP/UX 11i.

and the reason that Linux exists is that it answers a market need !
people are fed up of paying licence fee's for bloatware.

and as you say yourself a free product can give better performance
than its expensively licenced rivals !!

[michael newport]

$400 is less than we spend in a week for free softdrinks for our

employees. Get a life.

see a dentist !

[Billy Verreynne]

Why Oracle and not DB2? There are numerous sound technical reasons.

And this..

==
/home/billy/sqlplus dataware@whs
SQL*Plus: Release 9.2.0.5.0 - Production on Mon Nov 15 15:27:06 2004
Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved.
Enter password:

Connected to:
Oracle9i Enterprise Edition Release 9.2.0.4.0 - 64bit Production
With the Partitioning, Oracle Label Security, OLAP and Oracle Data
Mining options
JServer Release 9.2.0.4.0 - Production

SQLset timing on
SQLselect count(*) from x25_calls;

COUNT(*)
----------
672839836

Elapsed: 00:00:35.18

SQLexit
Disconnected from Oracle9i Enterprise Edition Release 9.2.0.4.0 -
64bit Production
With the Partitioning, Oracle Label Security, OLAP and Oracle Data
Mining options
JServer Release 9.2.0.4.0 - Production
==

Now anyone that have an idea what databases are about, will know what
a SELECT COUNT entails, I/O wise.. and how critical table and index
designs plays in optimising access and lowering I/O.

Can any other database, Open Source or commercial, come anywhere close
to this? I doubt it.

And no, this nothing to do with hardware. The above was run against an
old K-class HP-UX platform.

--
Billy

[Pete H]

what a dork...

Pete H
vs****@onwe.co.za (Billy Verreynne) wrote in message news:<1a**************************@posting.google. com>...

Why Oracle and not DB2? There are numerous sound technical reasons.

And this..

==
/home/billy/sqlplus dataware@whs
SQL*Plus: Release 9.2.0.5.0 - Production on Mon Nov 15 15:27:06 2004
Copyright (c) 1982, 2002, Oracle Corporation. All rights reserved.
Enter password:

Connected to:
Oracle9i Enterprise Edition Release 9.2.0.4.0 - 64bit Production
With the Partitioning, Oracle Label Security, OLAP and Oracle Data
Mining options
JServer Release 9.2.0.4.0 - Production

SQLset timing on
SQLselect count(*) from x25_calls;

COUNT(*)
----------
672839836

Elapsed: 00:00:35.18

SQLexit
Disconnected from Oracle9i Enterprise Edition Release 9.2.0.4.0 -
64bit Production
With the Partitioning, Oracle Label Security, OLAP and Oracle Data
Mining options
JServer Release 9.2.0.4.0 - Production
==

Now anyone that have an idea what databases are about, will know what
a SELECT COUNT entails, I/O wise.. and how critical table and index
designs plays in optimising access and lowering I/O.

Can any other database, Open Source or commercial, come anywhere close
to this? I doubt it.

And no, this nothing to do with hardware. The above was run against an
old K-class HP-UX platform.

[Billy Verreynne]

ph******@intellicare.com (Pete H) wrote:

what a dork...

And that is the best you can do Pete in response to a [SELECT COUNT]
on a VLT containing 672,839,836 rows that returns the answer in 35
seconds?

I've read Oracle being slammed for this and that and what not. So
instead of responding in kind, I simply show what Oracle is capable of
in the real world.

It is also not about counting rows in general. It is *what* it entails
(think I/O) and *how* it does it.

And the How It Is Done is what differentiate Oracle from others.
Inovative means of providing accurate and consistent answers - thus
enabling this very visible performance with a [SELECT COUNT]. And
this type of innovation and performance is across the board. Not just
with a [SELECT COUNT]. Though the latter tend to drive home the point
with an extra sharp and shiny edge.
--
Billy

[Lady Chatterly]

In article <63*************************@posting.google.commi************@yahoo.com (michael newport) wrote:

>
Serge,

would you like to see these other IBM products OpenSourced ?

I see what you mean.

>Regards
Michael Newport

Why are you so sure?

--
Lady Chatterly

"I don't know who she is. I doubt that its a bot. I have my guess as
to who it is. Regard the frequency of posts. What frequent poster is
missing? That Be Packing, is an old mind trick. Ignore it." -- Pip

[Jasper Scholten]

Carex,

Totally not supported but try:

./runInstaller - ignoreSysPrereqs

Ofcourse no support and not recommended at all, but propably will work.

Cheers,

Jahudi

carex wrote:

>
Hi,

A few years ago (2000 ?) I installed an oracle client on a linux server
(oracle8iR2 on Debian Woody) and everything went beautifully. (with
DBD::Oracle)

Now, I need an oracle client on my new laptop. (debian Sarge)

I tried oracle 10g instant client but whitout success. See below.

zorro@armada:~/tmpOra/instantclient10_1$ ls -al
total 85788
drwxr-xr-x 2 zorro zorro 4096 2005-05-24 17:11 .
drwxr-xr-x 3 zorro zorro 4096 2005-05-23 19:02 ..
-rwxr-xr-x 1 zorro zorro 14428 2005-05-24 17:11 a.out
-r--r--r-- 1 zorro zorro 1461081 2004-11-08 21:25 classes12.jar
-r--r--r-- 1 zorro zorro 1353 2004-11-08 21:25 glogin.sql
-rwxr-xr-x 1 zorro zorro 13495923 2004-11-08 21:25 libclntsh.so.10.1
-r-xr-xr-x 1 zorro zorro 2121849 2004-11-08 21:25 libnnz10.so
-rwxr-xr-x 1 zorro zorro 1229425 2004-11-08 21:51 libocci10_296.so.10.1
-rwxr-xr-x 1 zorro zorro 913575 2004-11-08 21:25 libocci.so.10.1
-rwxr-xr-x 1 zorro zorro 66159152 2004-11-08 21:25 libociei.so
-rwxr-xr-x 1 zorro zorro 96517 2004-11-08 21:25 libocijdbc10.so
-rwxr-xr-x 1 zorro zorro 760686 2004-11-08 21:25 libsqlplus.so
-r--r--r-- 1 zorro zorro 1397543 2004-11-08 21:25 ojdbc14.jar
-r--r--r-- 1 zorro zorro 21299 2004-11-08 21:25 README_IC.htm
-rwxr-xr-x 1 zorro zorro 14428 2004-11-08 21:25 sqlplus
zorro@armada:~/tmpOra/instantclient10_1$ ./sqlplus
./sqlplus: error while loading shared libraries: libsqlplus.so: cannot

open shared object file: No such file or directory

zorro@armada:~/tmpOra/instantclient10_1$

Then I tried to install the client (ship.client.lnx32.cpio)
But this seems not to be possibe on a Debian. Dee below

zorro@armada:~/tmpOra2/Disk1$ ls -al
total 36
drwxr-xr-x 6 zorro zorro 4096 2004-10-21 01:06 .
drwxr-xr-x 3 zorro zorro 4096 2005-05-24 15:37 ..
drwxrwxr-x 8 zorro zorro 4096 2004-10-21 20:40 doc
drwxr-xr-x 4 zorro zorro 4096 2004-10-20 07:56 install
drwxr-xr-x 2 zorro zorro 4096 2004-10-20 07:56 response
-rwxr-xr-x 1 zorro zorro 1487 2004-10-20 07:56 runInstaller
drwxr-xr-x 7 zorro zorro 4096 2004-11-21 22:25 stage
-rwxrwxr-x 1 zorro zorro 4408 2004-10-21 01:06 welcome.htm
zorro@armada:~/tmpOra2/Disk1$ ./runInstaller
Starting Oracle Universal Installer...

Checking installer requirements...

Checking operating system version: must be redhat-2.1, redhat-3, SuSE-9,

SuSE-8 or UnitedLinux-1.0

Failed <<<<

And now I do not know what to do.
I did also try to install the old oracle8iR2 but also without success.

So, has someone already successfully installed an oracle client in Debian
Sarge ??? (and DBD::Oracle)

If yes, could I have a few tips about this install ??

Thanks
carex