In the Cisco Observability Platform, the data model is entity-centric. This means that no incoming OpenTelemetry™ (OTel) observation is stored without a corresponding entity. A schema is a template with rules to derive the entity types based on the incoming OTel data. The entity types relate to each other through relations called associations and are modeled per namespace. Each entity type has associated data types (metadata)- metric types, event types, attributes, and associations.

Examples:

  • The entities are stored in the Topology data store. The metrics, events, logs, and traces (MELT) are stored in different stores, which are optimized for those data types.
  • The entity type Service exists in the schema, and the observed entity type Login Service exists in the Topology store. The list of all possible attribute keys is available in the schema, and their values are available in the MELT data store.
  • The entity type Service has a parent-child relation with the entity type Service Instance. This is a type of association called aggregation.

Schema Browser

Schema Browser is a tool that provides a graphical representation of the entity types and their associations in your entire stack. With the Schema Browser, you can: 

  • Browse and view the entity types in your schema
  • View the associations among various entity types across cross-functional domains (namespaces)
  • View the metadata such as attributes, metrics, and events of entity types
  • Convert the selected data into a query

To open Schema Browser, click Explore > Schema Browser.

Explore Entity Types and Their Associations

The Schema Browser has the  Explore and Focus areas, and the Schema pane.

The Explore area displays all the entity types available in your schema. The entity types are represented as colored bubbles and grouped as per their domains (namespace). Each color represents a specific domain such as Kubernetes, Infrastructure, and Application Performance Monitoring.

When you double click an entity type in the Explore area, the entity type moves to the Focus area and you can view the connections (associations) with other entity types across the domains in your full stack. You can double click any number of entity types and view their associations with other entity types. When the entity type is in the Focus area, you can double-click again to remove it from the Focus area.

For example, double click k8s:ingress to view its associations with various entity types across the domains.

If you want to explore more about the associated entity types such as cloud:load_balancer, double-click it to view its associations with other entity types. You get a graphical visualization of how different entity types are associated with each other across the domains.

Search an Entity Type

To search an entity type in your full stack:

  1. In the Find entity search box, enter the name of an entity that you want to view. If the entity type is available in a domain, the entity type and associated domain are highlighted.
  2. Click the domain to view the related entity type.

The following screenshot displays the entity type container across the domains:

View the Entity Type Metadata

By default, the Schema pane on the right pane displays the total count of the attributes, metrics, and events in your schema. You can filter the metadata by specific keywords.

In the Explore area, when you single click an entity type, the Schema pane displays the hierarchy (Supertypes and Direct subtypes), connections (Associated types), active instances, the metadata- attributes, metrics, and events of the entity type. For example, you can view the metadata for the entity type k8s:cluster.

Click Go to Summary to view the total count of the metadata.

Query the Metadata

Currently, you can query the data of a single entity type at a time. To query the data for an entity type:

  1. Single click the entity type in the Explore area. The Schema pane displays the metadata associated with the selected entity type.
  2. Click Query entity data on the Schema pane.
  3. Select the required metadata: attributes, metric types, and event types.
  4. Click Execute query to open the Query builder window.

Execute the Query 

The Query builder provides the interface to view and execute the query. You can also update the query in this window. Click Run query to execute the query block and fetch the observability data as a record and JSON output.

Alternatively, you can open the Query Builder from Explore > Query Builder.

If the query contains any errors, you get an error notification. View the error details to locate the error, check the error type, and how to rectify it. After you rectify the error, click Refresh query.

The query that you have executed is automatically saved in the Query builder. Click Query History to open the History tab and view your previous queries. To execute a previous query, click Run Query.

Query builder also provides you with a few predefined queries. Click Query templates to open the Templates tab and view the predefined queries. You can also edit and update these predefined queries and execute them as required.

Currently, these queries are created for Kubernetes.

For more information about Query Builder, see Explore Any Data.