How Service Validation Works

Building Service Models › How Service Validation Works

How Service Validation Works

Each step in the service validation process tests one aspect of the service topology. The tests target the structure of the service to verify that a service model is valid and will not cause impact propagation calculation errors.

The Service Modeler performs validation automatically before each save attempt. You can also validate on demand by clicking Perform Service Validation .

Each validation test is assigned a rank, which determines its impact on the subsequent service save process. The ranks and their effect are as follows:

UNKNOWN: Indicates that the test cannot determine the effect on the service.
LOW: Indicates that the model is inefficient, is missing items, or could use cosmetic improvements.
MEDIUM: Indicates that the model structure may affect the operation of the state engine.
HIGH: Indicates that the service cannot be put online in its current structure.
CRITICAL: Indicates that the service cannot be saved.

The tests run in the following order:

CRITICAL: The topology graph must be connected. The topology must not contain unconnected subgraphs; it must be possible to move between any nodes of the topology using associations, or all CIs must have relationships with other CIs.
HIGH: The topology graph must not have multiple roots. The state or impact of all CIs should propagate to one node or CI.
CRITICAL: The root node or CI must not have dependents. The root node is the final target for state or impact propagation.
HIGH: The topology graph (or subgraphs) must not have cycles formed by relationships. State propagation through relationships must not form loops.
CRITICAL: Component services of the tested service must not have antecedent CIs. Specifically, no antecedents are prescribed by the tested service's topology; the component services may be complex.
LOW: All graph nodes must have labels.

Results display in the Service Validation Results dialog if errors occurred. The table in this dialog has the following columns:

Rank: Defines the error rank.
Type: Defines the error type. Types include EDGE and PATH.
Root: Defines whether the problem affects the root service CI.
Object: Defines the affected parent object.
Description: Describes the nature of the problem.

Click each result for the Modeler to highlight the objects and relationship involved in the error. Use this visual information and the Description to resolve problems before saving the service model.