Re: Minimalist ORACLE Installation

"Hans Forbrich" <forbrich_at_yahoo.net> wrote in message news:51yrc.6185$SQ2.1489_at_edtnps89...
> Sarah Tanembaum wrote:
>
> >
> > Some of my colleaque to look into the opensource database such as MySQL
> > and/or PostgreSQL since they are easy to install and require virtually
> > very little resources as compare to MS SQL*Server, Sybase, DB2, and
> > especially Oracle. Is it true?
> >
> > The reason is that I have a limited diskspace and memory.
> >
> > Thanks
>
> The smallest 'supported' footprint for Oracle that meets your criteria is
> Oracle Standard Edition (or Standard Edition ONE if your machine only
> supports 1 or 2 CPU).
>
> You can theoretically create a custom install smaller than that, but you
> have to know exactly what you are doing. (And it probably won't be
> supported.)
>
> The Standard Edition {ONE] has most, if not all, of the features that
> typical PostgreSQL, MySQL, MS Access and SQL Server shops are looking for.
> (This is NOT a challenge for a feature battle - just a stanement of what
> the developers think they want, in my experience.)

I think, most small, medium, and large corporation can do away with the above.
Just as computer hardware, most user buy the most expensive one to get the most features but they are using it just for wordprocessing, reading email, and
play games. Many, if not, most medium-big size corporation wasting their resources for something that they do not actually needs it.

>
> There are dozens of reasons why MySQL and PostgreSQL have a smaller
> footprint. As of yet, the commercial products still have an edge in
> reliability, scalability, and other capabilities.

But how much more reliability, scalability, and other features they have as compared to PostgreSQL for good size shop or MySQL for small shop? For sometimes I was led to believe that more expensive is better, and was proven wrong.

>
> Oracle, for example, provides intrinsic support for several additional
> datatypes AND their manipulation. While PostgreSQL and MySQL support many
> of the datatypes, the manipulation is in some areas still a ways off. As
a
> result, you end up growing the effective footprint by needing to add extra
> software.

But the beauty of opensource, it will not cost you anything to acquire an additional
software. On example is Apache! It is one of the best, perhaps the best Web Apps
Server around. It has many hooks too many things in which many commercial web server does not have. Apache becoming the pioner in many areas where other
commercial tend to follow.

>
> The kicker is that the additional things are preintegrated and tested with
> the database, whereas many OpenSource developers end up adding or
> integrating the capability manually. Some examnples, in Oracle's case:
> Apache-based HTTP listener, PERL, direct interface from Apache to database
> via mod_PLSQL, Java and J2EE, a command line interface to the data,
message
> & message queueing, email inteface, direct HTML capability, workflow, a
> text/document index and search mechanism, geospatial manipulation, XML as
a
> data type that can be joined with tables.

I think there are 2 camp of design method you mention above: modular and all-in-one.
As most big software house, they tend to be so bloated and a bit rigid, where the open
source tend to be modular, not by design, but by its nature where it is very collaborative
development.

There are good and bad being bloated and modulars. One is interopability amongst
modules, but at the same token, maintainability has perhaps inverse relationship.
I think it is easier to maintain a modular approach than all-in-one approach. That
said, it will also affect the cost of development.

Modularity will create more competition which is good for consumer where they can
easily pick and choose what they need the most, in which otherwise with the bloated
apps, they pretty much stuck with what they have in that one apps. Also, with modular
approach, they can purchase what they need not and add another later, but you can
do that with the big apps.

>
> (Similar statements can be made for the other products.)
>
> The counter argument is generally "I want to pick the version levels of
the
> add-ons". Which is fine if you want to spend the time and effort
> supporting the required combination.
>
> (One other _major_ difference is that Oracle uses a SCHEMA in a manner
> similar to other products' DATABASE. Many developers get this confused
and
> create many Oracle Databases when they really should have one database
that
> contains many schemas. That frequently results in a footprint that is
MUCH
> larger than necessary.)
>
> Don't get me wrong - I do like and use Open Source. I just believe in
> picking the right tool for the job, and understand WHY it's the right
tool.
> Many developers snub Oracle simply because they do not know what it is
> capable of doing. Thus they end up reinventing the wheel - which may keep
> the initial cost low but tends to increase the long term operating cost.

Opensource does cut cost substantially if you know what you are looking for. Perhaps the benefit of commercial software are good for those who do not know
for sure exactly what they are looking for. It is easier to pick the most features and
figure out later what feature are needed, isn't it?

>
> HTH
> /Hans
>
>
Received on Sat May 22 2004 - 12:36:37 CDT