[thelist] Relational DB Help

David Kutcher david_kutcher at hotmail.com
Sat Feb 23 15:17:01 CST 2002


Nick -

> > 1. less strain on the database to do queries on large amounts of data
> That strain will be picked up by something else though. There are also
> some pretty nifty tools for such circumstances.

Yes, but this would allow you to separate the huge amounts of text from the
database allowing for faster page creation, and would also allow you to
separate content queries and the related strain onto a separate system if
necessary.

> > 2. ability to use specialized tools to index and query content
>
> Like what?

I'm personally a huge fan of ht://Dig .  Check out the features.
http://www.htdig.org/

> > Also, lets say we're using MySQL and not a super-scalable, massively
> > relational database like Oracle with great indexing functionality...
MySQL
> > would choke.
>
> I don't see why?

There are inherent differences between Oracle and MySQL that would be an
essay and a thread unto itself.  For most projects, Oracle is an overkill
and MySQL is the way to go.  But there are definitely situations where
Oracle outperforms MySQL.  A large database (100,000+ records) of articles,
each indexable with a text blob of 100,000 characters would be a hefty MySQL
database.  Stored procedures and advanced indexing capabilities would be
required to run effectively.  MySQL has neither... nor is it's MySQL's
strong-point to provide these capabilities.  PostGres... possibly... if you
want to stay free/open source.   But not MySQL.

> > running a query for "the" in a potentially huge database because they
would
> > be separate process/applications.
>
> Yeah but the trick here is not to allow a user to search for 'the' or
> 'a' etc.

That's a separate issue and has to do with removing these from the client
side before performing the query.

> > but it might make sense here unless
> > he's using a big relational database on a real server system.
>
> If he had a dedicated MySQL server I don't think he'd have a problem
> although I wouldnt attempt it. Does he have /that/ much content?

Unknown, which is why I'm trying to explain the pro/con of both approaches
as opposed to advocating just storing the articles in a text field in the
database.

David
www.confluentforms.com




More information about the thelist mailing list