Skip to main content

Observability Feasibility Study (Grafana Cloud)

Objective: Determine the viable path to a "Single Pane of Glass" dashboard for Pulseway, RocketCyber, and Bitdefender metrics.

1. Pulseway Integration (T018)​

Findings:

  • Native Exporter: No official Prometheus exporter exists.
  • API Strategy: The Pulseway REST API provides endpoints for Systems (status, CPU, RAM) and Notifications.
  • Recommendation: Use Grafana Infinity Data Source.
    • Config: HTTP URL https://api.pulseway.com/v2/systems, Auth Bearer Token.
    • Visualization: Table/Stat panels parsing JSON response data.systems.
    • Limitations: API rate limits need to be checked for high-frequency polling (e.g., every 1m).

2. RocketCyber Integration (T019)​

Findings:

  • Loki Webhook: RocketCyber supports generic webhooks for incidents.
  • Strategy: Send webhooks to an intermediate endpoint (e.g., n8n or AWS Lambda) that formats the JSON into a Loki log stream, OR use Odoo as the log aggregator (via the Helpdesk integration) and visualize Odoo data in Grafana (PostgreSQL datasource).
  • Recommendation: Indirect Integration. Visualize the Tickets created in Odoo (Source: RocketCyber) rather than raw RocketCyber streams, to ensure "Actionable" dashboarding.

3. Bitdefender GravityZone Integration (T020)​

Findings:

  • API: GravityZone has a reporting API (/stats/quarantine, /network/incidents).
  • Strategy: Similar to Pulseway, use Grafana Infinity to poll the API.
  • Recommendation: Grafana Infinity.
    • Endpoint: /api/v1.0/jsonrpc/network/incidents.
    • Auth: API Key (Base64 encoded).

"The Hybrid Dashboard"

  1. Infrastructure Health (Pulseway):
    • Source: Grafana Infinity Plugin (Direct API Polling).
    • Panels: Fleet Status (Online/Offline), CPU Heatmap, Low Disk Space Table.
  2. Security Incidents (RocketCyber/Bitdefender):
    • Source: Odoo PostgreSQL (via SQL Datasource).
    • Reasoning: Since all critical alerts trigger Odoo Tickets (as per Phase 3), querying the Odoo Ticket table for tag='Security' is more reliable and reflects "Work to be done".
    • Panels: Open Security Tickets Count, Mean Time to Resolve (MTTR).

Conclusion: Do not try to replicate every raw metric. Visualize Health (via Infinity) and Action Items (via Odoo SQL).