Which database is the most scalable – MySQL, SQL, Oracle, something else?
Which of the above can handle the most queries quickly and efficiently?
Does the programming language affect the speed of the query? If so, what is the fastest?
Is there a size limit to a single database on any of the database management systems? (MySQL, SQL, Oracle, etc?)
If you were designing a website with millions of database records, queries, and page views per day, what would you choose? What if money was not an issue - unlimited funding? Would your answer change?

Originally Posted by bobena

Which database is the most scalable – MySQL, SQL, Oracle, something else?

Probably Oracle, with SQL Server as a close #2, and MySQL as a distant #5. PostgreSQL, SQL Lite, and others come in between.

Originally Posted by bobena

Which of the above can handle the most queries quickly and efficiently?

It depends somewhat on the specific queries and data, of course. But again, it's Oracle and SQL Server.

Originally Posted by bobena

Does the programming language affect the speed of the query? If so, what is the fastest?

Absolutely! Good coding is essential. In any language.

Originally Posted by bobena

Is there a size limit to a single database on any of the database management systems? (MySQL, SQL, Oracle, etc?)

Yes. Thousands of terabytes. In most real world scenarios you'll be limited by the amount of money you'll spend on drives and on architecture, rather than by the amount of storage a computer can address.

Originally Posted by bobena

If you were designing a website with millions of database records, queries, and page views per day, what would you choose? What if money was not an issue - unlimited funding? Would your answer change?

I would choose SQL Server (2008, Enterprise Edition) because it's among the best, and, more importantly, because I have the most detailed knowledge of it, allowing me to build out competent architecture to support whatever I was trying to do.

Now some questions

1. Are you building a transactional, or reporting system?
2. How is your data structured?
3. What types of queries will be asked, by how many people, at any given time?

The code you put into the database (ie, what you ask it to do) is far more important than whether you choose SQL Server or Oracle. Both are up to the job of running good code against hundreds of millions of rows. Both will easily choke on bad code with very small data sets. Unless you know the difference between an index scan and a bookmark lookup, it won't matter which RDBMS you choose.

'm wondering how to design that part, should I store all information
on the web host only ? miror that DB every evening on some local DB
server to be able to use the data without eating up lots of bandwith ?
separate the database in 2 parts ? how to sync and assure integrity
then ? having a local DB will also mean the company will have to pay a
licence for the DBMS ... >
I'd use Sybase, and I'd use replication server to move data from the
production server (in co-location facility) to the reporting server (in
the office).
>
But you can do similar things with Oracle, MS-SQL, etc., although I'd
personally stay away from MS-SQL due to its being Microsoft-centric (i.e.
can't scale beyond Intel/Windows type machines)
>
Michael
Dual DB's aren't really necessary.

Points in favor of 2 DB's
*its standard approach Production and Development DB's
*large reports don't hog resources from live interactions.
*new developments/data are staged in development first, in a secure,
isolated environment.

points in favor of 1 DB
*if its accessable from the WEB then developers can work there too.
(schema isolation should be as good as DB isolation)
*web access means you know when the site goes down.


I've seen both.