> i ask about the stability of the database for big system database does
> postgresql stable like oracle and when there is a power faliure oracle
[quoted text clipped - 3 lines]
> also can you help in making a difference table between both dbs
> thanx
Define "big system." Because by Oracle definition of what constitutes
a large or very large database postgresql isn't even in the running.
The lack of instrumentation, alone, would make such a beast untunable.

Signature
Daniel A. Morgan
University of Washington
damorgan@x.washington.edu
(replace 'x' with 'u' to respond)
>i ask about the stability of the database for big system database does
>postgresql stable like oracle and when there is a power faliure oracle
>solved this problem what about this problem (power faliure)in
>postgresql db ?
>also can you help in making a difference table between both dbs
You might want to try asking this question on the PostgreSQL lists -
they will try to answer honestly. As Daniel says, what is a "big
system"?
Paul...
>thanx

Signature
plinehan __at__ yahoo __dot__ __com__
XP Pro, SP 2,
if Oracle group then
db := Oracle 9.2.0.1.0;
else
db := Interbase 6.0.2.0;
endif
As a courtesy to those who spend
time analyzing and attempting to
help, please do not top post.
DA Morgan - 19 Mar 2005 21:04 GMT
>>i ask about the stability of the database for big system database does
>>postgresql stable like oracle and when there is a power faliure oracle
[quoted text clipped - 10 lines]
>
>>thanx
A "big" system in the world of Oracle is one where the size is at or
approaching 10TB.

Signature
Daniel A. Morgan
University of Washington
damorgan@x.washington.edu
(replace 'x' with 'u' to respond)
> i ask about the stability of the database for big system database does
> postgresql stable like oracle and when there is a power faliure oracle
[quoted text clipped - 3 lines]
> also can you help in making a difference table between both dbs
> thanx
By Philip Howard, Bloor Research
Published Thursday 24th March 2005 15:23 GMT
I have espoused the cause of Ingres for some time and, especially, I
have commended Computer Associates' decision to take the Ingres
database into the open source community.
Now, I dare say that at some point CA will tell me all about its "open
source challenge" and how successful that has been in developing new
facilities for Ingres. And they will also tell me about how many
downloads there have been (though I am cynical about the utility of
such a figure). But this article is about one company, Datallegro,
which has adopted Ingres as its database.
Datallegro is a vendor in the data warehouse space that develops and
markets an appliance-based solution. That is, a combination of
hardware and software similar in concept (if not in method) to that
which underpinned the original development of Teradata. The company is
headed up by Stuart Frost, who was the founder and CEO of Select
Software and subsequently chairman of Avellino.
Datallegro is an open source based product to the extent that it runs
on Linux and uses Ingres as its database. However, it wasn't always
like that. Originally, it was developed using MySQL. For performance
reasons, this was quite quickly replaced with PostgreSQL. Bear in mind
that at this time, Ingres was not available as an open source product.
When it became available, Datallegro evaluated Ingres and rapidly made
the decision to jump ship from PostgreSQL to Ingres (not too difficult,
as they share common origins). This says quite a lot about the relative
robustness, performance, scalability, and so on of MySQL versus
PostgreSQL versus Ingres.
You would not normally mention Ingres and data warehousing in the same
breath, yet Datallegro has tuned the product up to the extent that it
is offering one of the best (the best if you listen to Datallegro, but
I'll hedge my bets until some benchmarks have been published)
price/performance offerings on the market.
The key to this is that the core functionality and the architecture
were already in place within Ingres (not to mention flexibility), on
which Datallegro has been able to build. What is more, it is clear from
talking to Datallegro, just how committed CA is to open source.
Stuart Frost is extremely complimentary about the help that CA has
provided throughout the company's developments, describing it in
superlative terms. That is good to hear and should be music to the ears
of any other companies thinking of taking on major developments in
conjunction with CA and Ingres. ®
DA Morgan - 26 Mar 2005 18:08 GMT
> You would not normally mention Ingres and data warehousing in the same
> breath, yet Datallegro has tuned the product up to the extent that it
> is offering one of the best (the best if you listen to Datallegro, but
> I'll hedge my bets until some benchmarks have been published)
> price/performance offerings on the market.
price/performance is one very small measure of the value of a database
product.
What about security? How secure is a product when any 16 year old can
read the source code?
What about stability? Who do you call for support when it crashes like
an airplane without fuel and your customer's (internal and/or external)
expect it to be up 7x24x365?
What about future support? Are there any guarantees that one is not
investing heavily in something that will never mature beyond where it
currently is? The rest of the industry evolves and Ingres goes where?
What about resources such as books? magazines? classes? user groups?
I see Ingres as being close to death no matter how good the product
once was and nothing short of a robust organization that takes the
task of providing full support, such as RedHat with Linux, will revive
the patient.

Signature
Daniel A. Morgan
University of Washington
damorgan@x.washington.edu
(replace 'x' with 'u' to respond)
Lennart Jonsson - 28 Mar 2005 16:21 GMT
>What about security? How secure is a product when any 16 year old can
>read the source code?
Closed source as a security mechanism?
/Lennart
Leif B. Kristensen - 28 Mar 2005 17:45 GMT
Lennart Jonsson skrev:
>>What about security? How secure is a product when any 16 year old can
>>read the source code?
>
> Closed source as a security mechanism?
It's called "security through obscurity". Security experts generally
agree that it's a bad idea, though.

Signature
Leif Biberg Kristensen
http://solumslekt.org/
DA Morgan - 28 Mar 2005 19:17 GMT
> Lennart Jonsson skrev:
>
[quoted text clipped - 5 lines]
> It's called "security through obscurity". Security experts generally
> agree that it's a bad idea, though.
Security experts also agree that it is harder to break in when you don't
know what is protecting the door.
Think of the difficulty any cracker would have breaking into DB2 on the
mainframe. They don't have access to the hardware, the operating system,
or the software. Makes the job not impossible ... just more difficult.

Signature
Daniel A. Morgan
University of Washington
damorgan@x.washington.edu
(replace 'x' with 'u' to respond)
DA Morgan - 28 Mar 2005 19:16 GMT
>>What about security? How secure is a product when any 16 year old can
>>read the source code?
>
> Closed source as a security mechanism?
>
> /Lennart
One of many.

Signature
Daniel A. Morgan
University of Washington
damorgan@x.washington.edu
(replace 'x' with 'u' to respond)
--CELKO-- - 28 Mar 2005 19:29 GMT
>> I see Ingres as being close to death no matter how good the product
once was and nothing short of a robust organization that takes the task
of providing full support, such as RedHat with Linux, will revive the
patient. <<
Would Computer Associates count as a "robust organization" with a few
billion dollars behind it? They use Ingres internally in their
software, they have interfaces for it to their packages, etc.
Likewise, how many mainframe and mid-size packages have Informix 4GL,
Progress or Pervasive RDBMS engines under the covers? These guys are
the "professional tool" rather then the "DIY" tool for development.
The Ingres optimizer is still one of the best on the market. Oracle
and other RDBMS products are still tryignto play catch-up with Ingres,
which has been ahead of the game from the start. Stonebreaker really
worked those grad students at Berkeley to death :)
Security is not a matter of source code; it is a matter of usage.
Example: I know the algorithms for RSA encryption, but it is still
secure.
Stability is actually pretty good if CA will be the gatekeeper on the
language. If you look at the stats, the downtime for Ingres is damn
low.
>> What about resources such as books? magazines? classes? user groups?
<<
Those are problems. Books? The language is pretty much Standard
SQL-92, so it is quick and easy to learn, if you know another SQL that
is also close close to Standard SQL-92. Hell, maybe I will write
"Ingres SQL for Smarties" since I keep getting grief from my friends
who are "Johnny One-Product" programmers. Seriously, the last two
reviewers I had for the third editon complained that I did not have
enough MySQL or DB2 in a book on Standard SQL.
I am hoping that CA will make classes available for a cheap price, post
on-line tutorials, etc.. User groups are there but that takes time to
build them to a size that they can have meetings.
I am not sure it is dead yet. I would say it is a "cult classic" that
might be about to hit "regualr theater release" if the promotions are
done right.
The quality of the product is not the question; the question is will CA
support it when it is Open Source? I don't know if they will.