Re: Oracle vs. SQL Server 2

[irt]

Could someone please explain the basic differences between Oracle and SQL server relating to: (a) cost differences (b) ease of using Microsoft Access 2000 forms, queries, and reports with SQL server vs. the same with Oracle (c) which would be more suitable as a long-term upgrade from an Access 2000 database of about 175,000 records, growing at the rate of about 5,000 records per year. Thank you very much for any information you could give me.

 

[jimmythegeek]

You will find those who argue both side of the coin, but in my opinion, unless you are a large company with lots of money, your only option is SQL Server. Oracle claims to be the best, and may be, but they are too expensive for almost all small companies.

SQL Server on the other hand is much more affordable (retails for around $1,100.00 depending on version), and licenses are cheaper, and you don't have to have anything but a Windows 2000 server to run it.

I have used SQL Server as back-end to Access front-ends (unbound forms using ADO for data exchange and manipulation) multiple times with great success. The performance is great.

Also, I don't know how easy Oracle is to use, but SQL Server is pretty straight forward with an intuitive user interface.

I am sure others can argue the case for Oracle, but based on the fact that both will do what you need, the cost leans heavily toward SQL Server. Jim Lunde
compugeeks@hotmail.com
We all agree that your theory is crazy, but is it crazy enough?

 

[JimHorton]

irt,
I have to agree about the cost side of things, but the question is:
Do you want/need a BMW 750i (Oracle)?
or
Can you get by with a Chevy Cavalier (SQL7)?

With only 175,000 records and a small growth rate, and what appears to be a need for quick learning curve, low cost and not much other functionality/scalability, the cavalier will do.

I've used both, though Oracle is by far my most common back-end server. If I were doing anything mission-critical, I would never consider sql7 at this time. In my opinion, sql7 was another rush-to-the-market-a-pale-copy-of-the-market-leader project by Microsoft. Maybe in a few years sql7 will reach the flexibility and scalability of Oracle, but then, Oracle is moving forward at the same time.

Note to Larry Ellison--You're no better than Bill Gates when it comes to greed--the difference is that you happen to have a truly high-quality product, but....we're in a recession, and if you want to remain the market leader you should listen to posts like this and SHARPEN YOUR PENCIL!
--Jim

 

[irt]

Mr. Horton:

Thanks for the timely & informative response, and I'll consider it. Next question: Would learning all about SQL7 now help me shorten the learning curve for Oracle on my own later without my having to take another class, if my company were to decide to buy Oracle instead? To put it another way: if I learn SQL7 and then someone were to clandestinely "switch" me to Oracle, how much difficulty would I encounter when trying to do the things I got used to doing in SQL7?

Appreciate all the help I'm getting here, by the way.

 

[jimmythegeek]

Jim,

I ran SQL Server 7 and then upgraded to 2000, and I don't know that I would go with the Chevy Cavalier analogy. It has run several intranet business critical applications 24/7 since we purchased it over 1 1/2 years ago without so much as a hickup.

I am sure Oracle is a wonderful product, I would never argue that, but I believe SQL Server is not given the credit it is do.

In a conversation with a previous manager on whether their company would move to Oracle or SQL Server, he told me Oracle told them it would be between $150,000 and $200,000 to get their system up and running when all was said and done, and SQL Server Enterprise version would have cast under $20,000. So I agree with your note to Mr. Ellis in that how can ANY small business even DREAM of using Oracle when you have to pay that kind of money.

I work for a world-wide financial orginazation (over 20,000 employees) that has one of the largest Oracle databases in the world, and when we have to get a new Oracle server we start about $75,000 and go up. We have to have "Oracle DBA's to run them, and anything we want done in the database must go through them. We set up our Windows 2000/SQL Server/IIS machine in 1 day, and I control the entire server myself.

So in these circumstances, I will go with the "Cavalier" any time. Jim Lunde
compugeeks@hotmail.com
We all agree that your theory is crazy, but is it crazy enough?

 

[JimHorton]

irt,
Learning any SQL rdbms, be it Oracle, Sql7, Informix, Mysql, Sybase, etc, will be of help since for the most part, the sql lanquage is universal. There will be minor syntax differences, data type differences, function differences, but again, these are minor.

As far as the hypothetical clandestine switch, you'd see more of a problem if you were a dba as opposed to a developer. For example, I have Oracle and SQL7 servers running at home on the same box for my development work. I have a test front-end in vb, where I've created identical back-ends in Oracle and SQL7 for some simple performance benchmarking. For much of this test app (granted, it's a very simple front-end) I can simply switch the connection string and do nothing else, and I'm running the different backend--totally transparent. Yet building the two db's had much more differences--I couldn't create a build script from Oracle and use it directly in Sql7.

jimmytheGeek,
OK, a Caprice
But one of the problems is the classic blessing/curse of ease of use, Wizards, etc. I've seen large sql7 db's that were created very quickly. However, simple dba rules are totally ignored, such as indexes and table data residing not only on the same disk, but in the same file! The performance is of course nowhere near what it could be, though when compared to something like Access, where some of these db's had a previous life, the developer/dba sees a big gain and figures the problem is solved. And yes, in Oracle the same mistakes can be and are made, but Oracle forces more thought to be put into critical things such as the basic physical db properties.

I guess it's my personal preference of control--a bit old fashioned I know--I'd rather have a bit more work in the details and gain the detail level control. Ie, I still rarely use Wizards of any type in Access, and I know a guy sitting next to me can have a form created faster than I if he's using the wizard, but in the long run, I believe my form is more stable and useable, and total dev. time is less, because of all the annoying 'defaults' that the wizard creates, which cause problems down the road or have to be re-done anyway. And I still order the stick-shift transmission too
--Jim

 

[jimmythegeek]

Jim,

I completely agree. It is the the DBA's responsibility to see that the database is created correctly and efficiently, and I also do not use wizards in Access.

I would like to use Oracle more, but our DBA's have those things locked up so tight here, that it would be close to a special military operation to get anything done, or learn anything.

Besides, Microsoft says they are as good as Oracle, and who can question that? Jim Lunde
compugeeks@hotmail.com
We all agree that your theory is crazy, but is it crazy enough?

 

[Armadi]

HEllo
What advantages are there to migrate an app from Access to SQL server?
Thanks.

 

[jimmythegeek]

The two main advantages are performance and storage capabilities.

Access resides on a file server, which means the user's machine will do all of the processing while sending and retrieving excess data.

SQL Server is designed to handle all processing of requests, and only sends back results, thus improving performance. Besides the fact that a server will usually be bigger and better than user's machines.

Access can adequately handle (depending on the number of fields in a the tables) up to maybe a few hundred thousand records. Performance will definitely suffer however if you push Access to the limits (I actually saw an Access database with over 3 million records in it, and it took about 3 minutes just to move from one record to the other).

SQL Server is designed to handle millions of records without significant performance loss.

If you don't need to switch to SQL Server, then don't. It is more expensive and harder to administer. Access will work fine for smaller companies that don't have larger record counts. Make sure you weigh the costs of switching (learning curve, cost, administration, licensing, etc.), before you decide to do so.

I do believe if done correctly, a company can benefit significantly from switching. Jim Lunde
compugeeks@hotmail.com
We all agree that your theory is crazy, but is it crazy enough?