Filtering the synchronized components

All device nodes which match the filter created in Filtering the root device node are synchronized to the CMDB. After a node has passed the first filter, the synchronization system retrieves the source subgraph of all the relevant related nodes, known as its components. Before the source subgraph is transformed into the target subgraph, nodes can be excluded from the source subgraph using the Component Filter. If nodes are excluded from the subgraph, the transformation into the CMDB model proceeds as if the excluded nodes did not exist in the Discovery data store.

To select nodes to filter

1. Click the Component Filter tab of the CMDB Sync page.
   The left pane shows a tree of all the related node kinds in the source subgraph. To exclude all the nodes from a particular branch of the tree, de-select the check box.
2. To filter nodes in the subgraph based on their attribute values, click the relevant entry in the left pane.
   The right pane now shows the attributes defined for the chosen node kind.
3. Click an attribute name to add a filtering condition for that attribute.
   After a filter has been defined for a particular part of the subgraph, only those nodes that match the filter are included in the source subgraph.

Currently the labels in the Component Filter show the node type and a brief description. For generic nodes such as Details, the description is insufficiently detailed to identify which node is selected. You need to check the CMDB after synchronization to ensure that the correct node is selected.

Example of creating a filter for synchronizing web servers

The following example shows how to use the Filter Builder to create a filter that will synchronize only those software instances representing web servers on those hosts previously identified as web servers.

1. From the left pane of the filter builder, select Software Instance.
   The right pane is populated with Software Instance attributes.
2. Click the Type attribute.
3. Enter "IIS" in the text entry field and leave the condition drop-down on "contains word".
4. Click the Type attribute.
5. Enter "Apache" in the text entry field and leave the condition drop-down on "contains word".
6. Change the Software Instance where... drop-down to "any".
7. Click Save.

With this filter in place, any Software Instances (SIs) that do not match the conditions are removed from the source subgraph. When the source subgraph is transformed into the target subgraph, it is as if the chosen SI nodes were the only ones present in the Discovery datastore.