• C++ Programming for Financial Engineering
    Highly recommended by thousands of MFE students. Covers essential C++ topics with applications to financial engineering. Learn more Join!
    Python for Finance with Intro to Data Science
    Gain practical understanding of Python to read, understand, and write professional Python code for your first day on the job. Learn more Join!
    An Intuition-Based Options Primer for FE
    Ideal for entry level positions interviews and graduate studies, specializing in options trading arbitrage and options valuation models. Learn more Join!

MSSQL vs Oracle SQL

Which one do you prefer and why?


  • Total voters
    9
  • Poll closed .
Joined
1/13/11
Messages
1,362
Points
93
You can also comment on management studios. Undoubtedly both have serious problems at least with current versions.
 
Overall performance. I attended the conference on Oracle's Exadata, Exalogic and Exalytics yesterday and they made an exaggerated impression about the data retrieval operation speed with oracle server compared with HP and IBM and blamed on SQL(Hardware and Software engineered to work together :)). Some type conversion is more efficient in MSSQL than in Oracle and vise versa. Which one do you use more often for your custom applications?
 
I find MS SQL quite a lot easier to use and in particular sucking data in is rather easier. A depressing % of my life has been coercing data into a DBMS, MS SQL reduces that pain.

Be aware that of all the systems vendors I have ever dealt with, Oracle are easily the most maliciously actively unpleasant.
If they can screw you through going over your head to senior management, they will do that as a reflex and will fuck you over on licensing at the first opportunity and if one doesn't come along they will create it.
 
I have worked with both (not recently though). I agree with Dominic about Oracle's salesforce, they will try to suck your blood from everything. However their database is beyond solid. I experienced a lot of pains with MSSQL including corruption. When that happened, it was beyond painful to fix. I have never had an issue with Oracle and we did all sort of crazy stuff.

One thing that will trip you. IIRC, Oracle SQL flavor, PL/SQL comes from PL/1 and it is somewhat unorthodox at times if you are not familiar with that style. Not that T-SQL is any better but it is simpler.
 
Tangential topic.:coffee:

I was tinkering with a test Oracle system one time, and found out that if you create two triggers that "listened" for any DDL statements (i.e. DROP,ALTER) targeted at any of the triggers and prevented them from happening, you can create a virtually indestructible "island" of triggers (provided the permissions aren't set).

Anyhow, if an admin does this by mistake to her oracle database, she pretty much needs to re-install it (not even rollbacks will save it), like I had to for mine.
 
Tangential topic.:coffee:

I was tinkering with a test Oracle system one time, and found out that if you create two triggers that "listened" for any DDL statements (i.e. DROP,ALTER) targeted at any of the triggers and prevented them from happening, you can create a virtually indestructible "island" of triggers (provided the permissions aren't set).

Anyhow, if an admin does this by mistake to her oracle database, she pretty much needs to re-install it (not even rollbacks will save it), like I had to for mine.
You have DBMS_CHANGE_NOTIFICATION package for that now.
 
PostgreSql is free. MySQL is free as well.
Never worked with MySQL, but Postgres is somewhat hard to maintain, but I agree its very good choice for open source database. BTW, I thought that MySQL is not exactly free...
 
Never worked with MySQL, but Postgres is somewhat hard to maintain,
How do you know this? How hard? Any experiences or comments? BTW, I've been running Postgres 9.0 currently and it seems as hard/easy as any other database.

I thought that MySQL is not exactly free...

MySQL is completely free. If you want paid support you can pay for it. Using the database is still free if you want to do it yourself.
 
For what purpose? It depends. Can't use it for data warehouse. Would someone really save up 20% and use MS product - maybe not. :)
SQL Server can be used for warehouse, http://www.microsoft.com/sqlserver/en/us/solutions-technologies/data-warehousing/fast-track.aspx

TerraServer is an example of a warehouse.
Oracle can cost 5 times more of what MS product gives; the performance of both is comparable while the pricing is significantly not:

First Link in the Result Set:
http://search.microsoft.com/results.aspx?form=MSHOME&mkt=en-us&setlang=en-us&q=Do Not Pay Too Much For Your Database Licenses
 
from the document:

Enterprise/
Data Center
  • Premium high availability
  • Scalability
  • High-end management tools
  • Enterprise security
  • No CPU and memory limit
  • Microsoft SQL Server Enterprise Edition and Data Center Edition
  • Oracle Enterprise Edition
  • DB2 Enterprise Edition
$27,000–$55,000
per CPU or per Core


Postgre and MySQL are FREE, FREE, FREE. Not a penny. BTW, you will need to pay for the Windows license as well if you want to use MSSQL Server. At least Oracle and DB2 let you run on free OS.
 
Each of these DBs has its own niche and a market. Oracle, as per comparison doc, can cost 110k while MS SQL can be at around 20k for a comparable set of features. As number of CPUs in a system grows, so does the overall price; a 6-core system based on Oracle will cost 570k while MS will be at around 137k. Some companies need features of expensive RDBMS and are willing to pay. For the rest of the word, free versions will do just fine. Fraction of Windows OS cost is nothing as compared to a solution cost based on Oracle.
 
Some companies need features of expensive RDBMS and are willing to pay.

Which features? I need some examples.

Most of the big (and famous) technology companies use free databases. So, I don't know which features "expensives" RDBMS(s) provide that are not needed by hi-tech companies.
 
Just a few I would give right now would be the replication/mirroring functionality (Merge), Reporting Services that allow building reporting solutions in a matter of hours with no extra cost, business intelligence studio to do the DTS and many other tasks (and one can be a business manager, not a programmer) and also the Integration Services. It is possible to lease the server and pay a small fee to use a full functional MS SQL. Also, complex solutions (app + db) could be maintained by only one person as opposed to having a dedicated DBA with the Oracle. Again, the same problem is possible to solve using all type of DBs; some come with the pre-built features that allow not reinventing the wheel. Also, Oracle and MS SQL have free versions that can be used by businesses.
 
Which features? I need some examples.

Most of the big (and famous) technology companies use free databases. So, I don't know which features "expensives" RDBMS(s) provide that are not needed by hi-tech companies.

BTW, who does actually pay 570k and 137k for a database and what are the features so deadly needed to pay such amount? Just curious...
 
That price is for a scalable multi CPU solution. One example: http://www.microsoft.com/casestudies/Case_Study_Detail.aspx?casestudyid=4000003470


"Data Tier

. MSC has some 22 terabytes of information stored on multiple instances of SQL Server across its global operations. The data tier has 1,934 tables and 4,959 stored procedures. The system supports more than 500 million database transactions per day. The data tier is hosted on 18 Unisys ES7000/600 and ES7000/520 servers, each with eight Intel Xeon Dual-Core processors and 32 gigabytes (GB) of RAM. The servers are configured as two-node active/passive clusters. SQL Server 2008 handles about 396,907 transactions per minute, and about 210 billion transactions per year. Copies of the core data synchronize between the ES7000 server clusters at three global data centers. MSC synchronizes its databases across its three major data centers with Microsoft Message Queuing technology, which enables applications running at different times to communicate across heterogeneous networks and systems, providing reliable message delivery, efficient routing, security, and priority-based messaging."
 
Back
Top