Solving Operational Business Intelligence with InfoSphere Warehouse Advanced Edition

324 Solving Operational Business Intelligence with InfoSphere Warehouse Advanced Edition

򐂰 All table space backup images to be merged for each database partition in

sequence.

򐂰 New merged backup image is prepared on disk prior to being written back to

TSM.

Where the base backup image resides on TSM, it does not need to be staged to

local disk.

8.7.2 Install and configure db2mk in partitioned database

environment

Use the following guidelines when installing and configuring DB2 Merge Backup

in a partitioned database:

򐂰 Install DB2 Merge Backup on each data node in the partitioned database,

starting with the administration node and including each standby node.

Create a staging directory using the same directory name on each node. Only

one stage directory per data node is supported. DB2 Merge Backup defaults

to the /tmp file system if no staging directory is specified.

򐂰 db2mbk.cfg

is the configuration file used to determine default parameter

values.

The configuration file shown in Example 8-14 was used in our test

environment.

Example 8-14 Contents of db2mbk.cfg in /opt/IBM/DB2TOOLS/MergeBackup11/cfg

# db2mbk default configuration

db2instance=db2inst1

nbcpu=1

netservice=db2mbkdm11

stagedir=/stage/stage_db2merge/

Where:

–nbcpu

The nbcpu configuration parameter determines the maximum number of

processor threads that the DB2 Merge Backup process can use when

creating a merged backup image for a database partition. DB2 Merge

Backup is designed to process data as efficiently as possible using all

available resources. To avoid any unintended effect on resource usage,

set the nbcpu value to 1. Override where required through individual

control files.