Creating a filtered attribute group
Create a filtered attribute group (data set) by filtering rows of data from an existing attribute group. If an existing data set returns multiple rows, you can create a filtered group returning one row for use with IBM® Performance Management.
About this task
A filtered attribute group has the same columns as the source attribute group, but can exclude some of the rows. It uses a selection formula to determine which rows to include.
To provide status and summary information for Performance Management, you need to use a data set that returns a single row. For details, see Preparing the agent for IBM Performance Management. If the source information is in a data set that returns multiple rows, you can create a filtered attribute group that returns a single row.
For example, the process, Windows service, and command return code data sources provide information as rows in the single Availability data set. You can create a filtered attribute group, using the NAME field in the selection formula. The group includes status for the necessary application. Define it as returning one row. Then you can use this attribute group as the summary data set for Performance Management.
A filtered attribute group is also useful when a base data source query returns data that you prefer to divide into separate groups. Examples of such data sources are Windows Performance Monitor, SNMP and WMI.
Name Type Size Used Free
Memory MEM 8 4 4
Disk1 DISK 300 200 100
Disk2 DISK 500 100 400
This is a table that reports
on the storage that exists on the system and it includes both memory
and disk space. You might prefer to break down the table into memory
and disk as separate tables. You can break down the table by creating
two base attribute groups. Each of these base attribute groups collects
the same data and filters out the rows you do not want. However, that
is not the most efficient way to do things. Instead, you define one
base attribute group that returns both the memory and disk usage data
together. Next, define two filtered attribute groups. Each uses the
same base table as its source. One includes a filter where Type=="MEM" and
the other includes a filter where Type=="DISK".Name Type Size Used Free
Memory MEM 8 4 4
and where Type=="DISK",
the returned data is: Name Type Size Used Free
Disk1 DISK 300 200 100
Disk2 DISK 500 100 400