HP Large Objects System Management Manual

HP Large Objects System Management Manual – 543599.001
SQL catalog subvolume name
and as the first element of the
Buddy Set subvolume name. The
name must start with an alpha
letter for example, sys1.
catalogVolumeName none The name of the volume where
the SQL catalog is to be created.
Must be in the format \$xxxxxx
for example, \$data01
3.6.2 define.sh script
The define.sh script is loaded automatically by .profile and takes the catalog location
from the .profile settings. By default, the Large Object SQL tables are also created on
the SQL catalog subvolume. If desired the volume and subvolume can be amended to
relocate tables. Note that a table partitioning script is also provided to spread the load
across multiple volumes.
There are no other configurable variables used in this script.
3.6.3 createCatalog.sh script
The createCatalog.sh script does not accept any parameters. The location of the Large
Object SQL catalog is taken from:
Volume - catalogVolumeName from .profile
Subvolume - objectStoreNumber from .profile
Invoke the script from the OSS prompt by typing:
createCatalog.sh
3.6.4 createObjectStore.sh script
The createObjectStore.sh script accepts parameters interactively via the OSS
command prompt.
Invoke the script from the OSS prompt by typing:
createObjectStore.sh
Accepted Parameter Default Value Usage
Number of objects 1 Entered in whole numbers of millions,
this is a target for the expected
number of objects that are to be stored
in the Large Object system. The
figure is used to calculate SQL table
extent sizes.
Number of SQL table
partitions
1 Certain SQL tables can be partitioned
(see createPartition.sh for more
details). The number of partitions is
collected here as it has an impact on
Page 24 of 37