Using Aggregation Rules

You can define aggregation rules in the config.xml file to optimize the configuration declaration. When you run the set system searchindex MQL command, these config rules are injected into the CloudView Consolidation Server.

This task shows you how to:

Flatten Child Data in Parent Object
Aggregate Multiple Types and Relationships

Flatten Child Data in Parent Object

You may need to flatten the file content of child objects within a parent object, to search for the parent object using the content of the child objects. This procedure describes how to configure the 3DSpace Index to create this aggregation rule.

Edit the config.xml file.
Define an aggregation rule like:
```
<AGGREGATIONRULE botype="Parent Object" aggr="mxfileinfo" type="STRING" 
              name="mxfileinfo" includefile="true">
              <CONNECTION kind="from" relTypeTarget="Reference Document" />
              <CONNECTION kind="to" boTypeTarget="Child Object" />
              </AGGREGATIONRULE>
```
In this aggregation rule:
- The Parent Object BOType is associated with the Child Object BOType using the Reference Document relationship type.
- The file contents of child objects are flattened on parent objects. The includefile="true" property activates the flattening.
Run the set system searchindex MQL command.
The MQL command generates a Consolidation Server aggregation rule, based on the config.xml <AGGREGATIONRULE /> definition. The 3DSpace Index Crawler uses this aggregation rule to flatten file content.

The file content of child objects is now flattened within parent objects.

Note: The validate searchindex command now reports issues between the parent object and the linked object, in case of file indexing failures.

Aggregate Multiple Types and Relationships

You can simplify your config.xml file by aggregating multiple types and relationships in a single rule declaration.

In <AGGREGATIONRULE> nodes, both botype and relTypeTarget attributes accept multiple comma-separated objects or relationships.

Edit the config.xml file.
Define an aggregation rule like:
```
<AGGREGATIONRULE botype="T1,T2" aggr="PHYSICALID" type="IDENTIFIER" 
name="isRefinedInto" sixw="ds6w:refinedInto">
	<CONNECTION kind="from" relTypeTarget="C1,C2,C3" />
	<CONNECTION kind="to" boTypeTarget="T11,T12,T13" />
</AGGREGATIONRULE>
```
In this aggregation rule:
- PHYSICALID is a consolidation field created from the PHYSICALID business object attribute.
- isRefinedInto and ds6w:refinedInfo are two index fields visible in the Exalead CloudView Administration Console data model.
- IDENTIFIER is the field type of isRefinedInto and ds6w:refinedInfo.
- The aggregation rule generated in the Exalead CloudView Administration Console is called sxirule_ruleGen_1 under Consolidation/sxiconsolidation0/Aggregation processors.
Run the set system searchindex MQL command.
The MQL command generates a Consolidation Server aggregation rule, based on the config.xml <AGGREGATIONRULE /> definition. The 3DSpace Index Crawler uses this aggregation rule.

In our example, for all business objects of T1 and T2, the Consolidation Server applies this rule to:

Take the physicalid values of the objects connected via relationship type C1, C2, or C3; to business object types T11, T12, or T13.
And push the values to the isrefinedinto and ds6w:refinedInfo index properties.