Skip to main content

Armis Integration

ServiceRadar ingests Armis device intelligence to enrich inventory, surface unmanaged assets, and drive risk-based alerting. The integration runs through the Sync service and optionally uses the Faker generator for demos.

Prerequisites​

  • Armis API client credentials with read access to your tenant.
  • A dedicated Sync worker or Kubernetes deployment with outbound connectivity to the Armis API.
  • KV entries for API keys and pagination cursors; see the KV configuration guide for secure storage tips.

Enabling the Integration​

  1. Populate the Sync config with the Armis connector block:

    {
      "integrations": {
        "armis": {
          "api_url": "https://<tenant>.armis.com/api/v1",
          "client_id": "${kv:secrets/armis/client_id}",
          "client_secret": "${kv:secrets/armis/client_secret}",
          "page_size": 500
        }
      }
    }

  2. Deploy or restart the serviceradar-sync workload. Watch kubectl logs for armis_sync messages confirming pagination progress.

  3. Confirm imported devices in the registry via srql:
    in:devices source:armis sort:risk_score:desc limit:20.

Demo and Testing​

  • The Armis Faker synthesizes data sets for local or CI demos. Mount the serviceradar-faker-data PVC to persist generated inventories.
  • Resetting Faker? Follow the Agents runbook to truncate Proton tables and replay canonical devices.

Alerting and Enrichment​

  • Sync attaches Armis risk scores and device tags; combine them with SNMP status to prioritize outages.
  • Use the Service Port Map to overlay Armis risk data on topology diagrams.
  • Export curated Armis datasets to third-party tools through the MCP integration.

Troubleshooting​

  • Authentication failures usually mean expired client secrets—rotate them in KV and redeploy Sync.
  • Large tenants may hit rate limits; tune page_size and enable incremental sync by storing the last_seen cursor.
  • For ingestion gaps, consult the Troubleshooting Guide and cross-check Faker vs. production statistics.