Operations Guide
This section is for teams that consume the observability platform — NOC, SREs, and operations engineers.
First Response
flowchart TD
A[Alert or Reported Symptom] --> B[Capture Customer Region and Time]
B --> C{Telemetry Current?}
C -->|No| D[Investigate Inventory Discovery or Ingress]
C -->|Yes| E{Service Signal Degraded?}
E -->|Yes| F[Correlate Metrics Logs and Traps]
E -->|No| G[Compare Peer and Expand Time Range]
D --> H[Route to Observability Owner]
F --> I[Route to Device or Service Owner]
G --> F
- Record the customer, region, data center, symptom, and incident time range.
- Confirm the alert or report against current Grafana and/or Splunk evidence.
- Check data freshness; stale or missing telemetry can look healthy.
- Compare a healthy peer to determine whether the scope is one node, cluster, customer, region, or the observability platform.
- Correlate service metrics with logs or traps before assigning ownership.
- Include dashboard/search links, filters, timestamps, and target status in the handoff.
If the Prometheus target is absent or down, treat that as a visibility problem until service health is confirmed through another source.
Getting Access
| Platform | Guide |
|---|---|
| Grafana | How to request and configure Grafana access |
| Splunk | How to request and configure Splunk access |
Using the Tools
| Guide | Description |
|---|---|
| Using Grafana | Navigate dashboards, read alerts, create panels |
| Using Splunk | Search logs, use saved queries, create reports |
Resources
- FAQ — Frequently asked questions
- Best Practices — Operational best practices
- Troubleshooting — Pipeline-by-pipeline checks