[object Object]

Manual top-down service mapping interviews — sit with the app team, draw the boxes, never refresh — are the single most thankless task in ITOM. The cloud-native answer is tag-based: the app team owns tags on their resources, ServiceNow reads the tags, the map builds itself and stays current.

The contract is the tag schema

You do not own the tags. The app teams do. Your job is to publish the schema, validate compliance, and translate tags into CMDB structure.

Required tags on every cloud resource:
  app_id          (matches CMDB Business Application code)
  environment     (prod | stage | dev)
  service_tier    (web | app | data | cache | queue)
  owner_email     (for routing)
  cost_center

If a resource is missing any required tag, it does not get mapped.

The discovery flow

Cloud Discovery reads tags into cmdb_ci_* records. A scoped Service Mapping job groups them by app_id and creates the application service entry plus inferred relationships:

For each app_id:
  Create cmdb_ci_service_auto entry if not exists
  Map tier=web -> Used by tier=app
  Map tier=app -> Used by tier=data
  Map tier=app -> Used by tier=cache
  Map tier=app -> Used by tier=queue

The inferred relationships handle 90% of cases. The remaining 10% are documented exceptions.

Tag compliance is a published score

Build a Performance Analytics indicator: % cloud resources with all required tags. Publish per business unit. Make it part of the cloud governance scorecard. Compliance jumps fast when the number is visible to a VP.

The CMDB fed by tags vs interview

A tag-fed application service is current within a discovery cycle. An interview-fed one is current as of the last interview. There is no comparison.

Handle multi-tagged resources

A resource with two app_id tags should error, not silently pick one. The reconciliation script must flag and route to the owning team for cleanup.

When tags are not enough

For request flow within the application — which microservice calls which — tags do not help. Use observability data via the Cloud Observability connector. Trace data fed into Service Mapping creates the inner topology. Tags create the outer topology.

What to do this week

Publish the tag schema as an organizational standard, even if compliance starts at 30%. Build the compliance dashboard. Set a 90-day target of 80% compliance. The discovery pipeline can wait until compliance is real.

[object Object]
Share