3. Me
• Terry Bunio
• Data Base Administrator
– Oracle, SQL Server 6,6.5,7,2000,2005,2012, Informix,
ADABAS
• Sharepoint fan
• Data Modeler/Architect
– Investors Group, LPL Financial, Manitoba Blue Cross, Assante
Financial, CI Funds, Mackenzie Financial
– Normalized and Dimensional
• Agilist
– Innovation Gamer, Team Member, SQL Developer, Test writer,
Sticky Sticker, Project Manager, PMO on SAP Implementation
– Currently all of these all the same time on an Agile SAP
implementation for an Insurance company
4.
5. Protegra
• In Business 14+ years
• Winnipeg Company
• Agile and Lean key to our practices
• 60+ consultants
• Software Development and Business
Performance Consulting
• Microsoft Gold Partner
7. The Project
• Enterprise Data Warehouse project
• Their existing Data Warehouse was:
– A combination of application databases, reporting database, and
data warehouse
• With a hockey pool database to boot
– In Sybase
– Reports created in Impromptu 7.0
– They needed to get off of Impromptu because of licensing costs
• They had a history of using Open Source products
– Linux
– Jasper Reports
– Java
• Although they had a history of Open Source products
they were open to other options
8. The Objectives
• Implement lessons that have been learned over the past 11
years since the Data Warehouse has been in production.
• Eliminate many potentially unneeded reports that have been
created over the past 11 years and minimize the maintenance
effort of the Data Warehouse
• Implement a Data Dictionary for the Data Warehouse that will
enable individual business departments to be much more
efficient at creating self-service reports
• Eliminate redundant data from the Data Warehouse that has
caused misunderstandings and difficulties in creating reports
from the Data Warehouse
• Improve the design of the data structures in the Data
Warehouse to address inconsistencies and less than optimal
design.
• Possibly reduce the license cost of the Data Warehouse.
9. The Requirements
• The Data Warehouse serves up 100% operational reports
• There is no Big Data requirement
• Advanced DBMS and Reporting tools really aren’t needed
• Just require a platform for non-sexy textual, operational reports
• Not a huge amount of data either – 1 Terabyte
11. Database Investigation
• Due to the cost of DB2 and the operational requirements, the
following options were identified as candidates and evaluated:
• Sybase ASE Enterprise Edition – Current Data Warehouse
Technology
• Microsoft SQL Server 2012 Enterprise Edition
• Oracle Enterprise Edition with Partitioning option
• Open Source Community Edition
• Open Source Supported Edition
14. Cost Investigation Scenarios
• 2 Servers with 2 quad core CPUs on each Server – 16 cores in
total
• 2 Servers with 2 hex core CPUs on each Server – 32 cores in
total
• 2 Servers with 4 hex core CPUs on each Server – 64 cores in
total
15. Investigation Constraints
• This investigation did not include the functionality or cost of:
– Disaster Recover
– Hot or Warm Failover functionality
• This is because this was not a requirement of their current
environment
17. Open Source Community
versus Supported Editions
• Community Editions– As is. No support outside of general
community and message boards in case of issues
• Supported Editions – 3rd party provides additional support in cae
of questions or errors. And example of this is EnterpriseDB that
provides this type of support for PostgreSQL.
18. Open Source Community
versus Supported Editions
• No sane corporate enterprise would run solely a community
edition of a database. Maybe other frameworks, but not a
database.
• So for a corporate enterprise, we will not consider the
Community Edition of Open Source DBMS
22. Annual Estimated Costs
• License Support costs – These costs are the support and
maintenance costs for both the Production and Development
environments.
• Estimated Annual Maintenance Costs – It is estimated that
there will be additional ongoing effort to maintain the different
DBMS solutions as compared to the current Sybase solution.
This is due to the different complexities of the DBMS solution,
the available tools for the DBMS solution, and the available
support in the DBMS user communities.
28. Voted off the island
• Due to the large cost of Oracle, Oracle was included from the
next phase of investigation
29. Risks to Strategic Database
Selection
• A change in the strategic technologies of the corporation.
• A limitation in the current technology as compared to its
competitors.
• A requirement to re-sell the database asset to another
company and the current strategic Database technology is not
consistent across companies.
• A requirement to re-use (i.e. host services) the database asset
externally and the current strategic Database technology is not
consistent across companies.
30. Mitigation activities to
minimize Risk
• 100% usage of only ANSI-compliant SQL to eliminate the
requirement to convert SQL if a change in database technology
is required.
• 100% usage of only ANSI-compliant Database features to
eliminate the requirement to convert away from these features
if a change in database technology is required.
• Full source code control and versioning of all Database Data
Definition Language (DDL) statements.
• Minimal usage of Database triggers and stored procedures as
most databases have stored procedures and triggers written in
a proprietary language.
• Creation of a Database isolation layer using views that can help
to mitigate changes required to the database for the new
consumer.
40. Gartner Comparison
• The criteria listed have been compared to the Gartner criteria
recommended to evaluate Data Warehouse solutions. The
percentages are different due to the fact that 30% of our
evaluation contains factors that Gartner does not consider.
These factors are:
42. Gartner Comparison
• The two factors that differ greatly between the two ranking
methodologies are:
– Performance
– Administration and Manageability
• The only factor that is weighted more in the Gartner ranking
methodology is the Administration and Manageability factor.
This factor is addressed in two areas on our document, Cost
and Functionality. A material part of the cost factor takes into
consideration the additional administration and management
that would be required.
• With this in mind, it is our opinion that the Administration and
Manageability factors are being weighted consistently across
the two ranking methodologies.
• It is our opinion that performance is rated too low in the Gartner
criteria
61. Lessons Learned
• I undertook this investigation in the formal manner I did
because I was disappointed with the lack of impartial reviews in
the market with enough detail
• Gartner has high level reviews, but nothing at a practitioner
level
• Any reports at the practitioner level has been ‘tuned’ to promote
a certain DBMS
– MySQL reports using MyISAM engine
– Oracle and Sybase reports using SQL built to perform best
on their configurations
62. Lessons Learned
• Open Source DBMSs are not the bargain they initially appear
to be
• Know your requirements
• Do your own tests
• Read the fine print
• Oracle really is expensive
63. References
• The State of Data Warehousing in 2011 – Gartner
• Data Warehousing Trends for the CIO, 2011-2012 – Gartner
• Comparison of Data Warehousing DBMS Platforms – Illuminate
• Next Generation Data Warehouse Platforms – The Data Warehouse Institute (TDWI) – Phillip
Russom
• Magic Quadrant for Data Warehouse Database Management Systems – Gartner
• PostgreSQL versus MySQL - http://www-css.fnal.gov/dsg/external/freeware/pgsql-vs-mysql.html
• MySQL versus PostgreSQL - http://www.wikivs.com/wiki/MySQL_vs_PostgreSQL
• Why PostgreSQL instead of MySQL 2009 -
http://wiki.PostgreSQL.org/wiki/Why_PostgreSQL_Instead_of_MySQL_2009
• Open Source Feature Comparison Matrix -
http://www.devx.com/dbzone/Article/29480?trk=DXRSS_DB
• The Open Source Data Warehouse Revolution -
http://www.ittoday.info/Articles/Open_Source_Data_Warehouse.htm
• MySQL 5.6 Reference Manual - http://dev.mysql.com/doc/refman/5.6/en/index.html
• PostgreSQL 9.1 Reference Manual - http://www.PostgreSQL.org/docs/9.1/static/index.html
• PostgreSQL vs. MySQL vs. Commercial Databases: It's All About What You Need -
http://www.devx.com/dbzone/Article/20743/0/page/1
• Cross Compare of PostgreSQL 8.4, SQL Server 2008, MySQL 5.1 -
http://www.Postgresonline.com/journal/index.php?/archives/130-Cross-Compare-of-PostgreSQL-
8.4,-SQL-Server-2008,-MySQL-5.1.html
• MySQL versus PostgreSQL - http://www.mysqlvsPostgreSQL.com/
• PostgreSQL versus MySQL: which is better -
http://www.databasejournal.com/features/mysql/article.php/3288951/PostgreSQL-vs-MySQL-
Which-is-better.htm
• Comparison of Relational Database Management Systems -
http://en.wikipedia.org/wiki/Comparison_of_relational_database_management_systems