Advanced Enterprise Server Edition
Posted: Mon Jan 27, 2025 9:17 am
These techniques can accelerate complex queries on large tables by a factor of 10-20. Customers report accelerations of a factor of 10-20 - in extreme cases even up to a factor of 100 or more. BLU stands for the implementation of the latest Blink technologies in DB2 LUW. BLU is not a standalone product or a separate database engine, but an integral part of DB2 LUW. Column-oriented and row-oriented tables can therefore be stored in the same database and evaluated together. Due to this deep integration, the BLU functionality is in principle included in every DB2 edition and only needs to be activated using one of these license keys:
IBM DB2 BLU Acceleration In-Memory Offering
Add-on Package for Workgroup and Enterprise Edition
Further details on the BLU technology can be found in the Wikipedia entry [1] and in IBM's "rapid adoption guide" [2] .
OLTP problem – tackling OLTAP with BLU?
The problem described at the beginning - analytical evaluations on operational data - could also be addressed by using BLU and converting the existing row-oriented tables to column orientation. Especially since the db2convert command [3] is a tool that can brazil telegram screening perform this conversion during operation with minimal downtime of the table. However, column-oriented tables also have their disadvantages in a predominantly OLTP-oriented environment. For example, single record operations and queries whose result sets have many columns and/or rows cannot be implemented as efficiently with column-oriented tables as with conventional row-oriented tables. And there are a whole series of restrictions for column-oriented tables - such as no support for:
Databases with a character set other than UTF-8 (Unicode)
With the upcoming DB2 LUW version 11, code page 819 will probably also be supported.
Triggers as well as check and referential constraints.
IBM DB2 BLU Acceleration In-Memory Offering
Add-on Package for Workgroup and Enterprise Edition
Further details on the BLU technology can be found in the Wikipedia entry [1] and in IBM's "rapid adoption guide" [2] .
OLTP problem – tackling OLTAP with BLU?
The problem described at the beginning - analytical evaluations on operational data - could also be addressed by using BLU and converting the existing row-oriented tables to column orientation. Especially since the db2convert command [3] is a tool that can brazil telegram screening perform this conversion during operation with minimal downtime of the table. However, column-oriented tables also have their disadvantages in a predominantly OLTP-oriented environment. For example, single record operations and queries whose result sets have many columns and/or rows cannot be implemented as efficiently with column-oriented tables as with conventional row-oriented tables. And there are a whole series of restrictions for column-oriented tables - such as no support for:
Databases with a character set other than UTF-8 (Unicode)
With the upcoming DB2 LUW version 11, code page 819 will probably also be supported.
Triggers as well as check and referential constraints.