Cost/Benefit Comparison of DB2 10.5 and Oracle Offerings
May 27, 2014 2 Comments
Danny Arnold , Worldwide Competitive Enablement Team
As part of the IBM Information Management team, I’m often asked to describe the advantages of DB2 10.5 over the Oracle offerings (Oracle Database 12c, Oracle Exadata, and other products). There are many reasons to choose DB2 10.5 over Oracle Database from a business value standpoint including licensing costs, technology advantages in the areas of compression, continuous availability with DB2 pureScale, and of course, the latest innovation, BLU Acceleration. BLU Acceleration combines columnar, memory optimization, and other technologies to deliver fast analytic query results and excellent compression along with greatly reduced administration. However, from a client perspective, it is someone from IBM stating that DB2 10.5 delivers all of these things (and we are probably a little biased).
Therefore, it was nice to read the recently published report from International Technology Group (ITG) Cost/Benefit Case for IBM DB2 10.5 for High Performance Analytics and Transaction Processing Compared to Oracle Platforms
This report describes both the high performance analytics and transactional processing application areas and highlights the advantages of DB2 10.5 over the Oracle offerings. There are a number of key pieces of information that this report brings to light including:
- DB2 10.5 provides a lower total operating cost of ownership than Oracle
- 28% to 34% lower 3-year TCO for transactional processing
- 54% to 63% lower 3-year TCO for high performance analytic
- Faster deployment with an average deployment time of 57 days LESS than an Oracle based solution
- 5.3X better compression rates for DB2 over Oracle (12.6X for DB2 versus 7.3X for Oracle)
The ITG report provides details in the areas of technology differentiators for high performance analytics where Oracle Database does not use Massively Parallel Processing (MPP) or an in-memory RAM based approach like DB2 with BLU Acceleration or SAP HANA, but provides another level of processing within the Exadata Storage Servers. So there is no built-in capability within the Oracle Database to efficiently process analytics in a high performance manner. Instead, the client must purchase an Oracle Exadata engineered system to gain any performance advantages for an analytics workload. ITG continues in their analysis of the Oracle Exadata system for analytics by stating:
The hybrid design has two important implications:
1. The overall environment is complex – administrators, for example, must deal with partitioned Oracle databases, RAC and Exadata-specific hardware and software features
2. Use of system resources is inefficient. High levels of system overhead are generated.
Exadata may be characterized as a “brute force” design. Because systems must compensate for overhead, the considerable processing power offered by this platform does not translate directly into application level performance.
The ITG report wraps up its discussion of technology differentiators describing how DB2 with BLU Acceleration takes advantage of the multiple processor cores available in today’s Intel and Power environments along with the other BLU technology advantages to deliver an average of 31.6X better performance versus the smaller average performance gain of 5.5X experienced by Oracle Exadata clients.
During the ITG discussion of complexity between DB2 and Oracle within the report, the findings were even more favorable to DB2. Oracle Exadata administrators have to develop skills in system and storage to augment their Oracle Database DBA skills. The average Oracle Exadata administrative task breakdown is 60% database administration and 40% system and storage administration. In ITG discussions with clients, Oracle Exadata systems required an average of 0.8 FTE (full time equivalents) per Exadata system for administration versus the DB2 with BLU Acceleration average of 0.25 FTE. This complexity difference between the two environments was highlighted in the deployment time comparison, with an average deployment time of 38 days for DB2 with BLU Acceleration versus an average deployment time of 95 days for Oracle Exadata. An interesting side note to this deployment time difference is that most Oracle Exadata deployments (over 60%) were performed by clients that were already Oracle Database owners and experienced with the Oracle Database.
The ITG report covers packaging and pricing and provides many tables and graphs highlighting the differences between DB2 10.5 and Oracle. If you are interested in learning more details about DB2 10.5 and its cost benefits over Oracle, I urge you to download and read the ITG report.