Skip to content
advanced ⏱️ 13 min read

Operations & Monitoring

Monitor platform health on skillSYMS using the ops cockpit: live signals, incidents, background jobs, exports, integration health, and the activity log.

For: Administrators

Summary

This guide covers the operational tooling platform administrators use to keep skillSYMS healthy. You’ll learn how to read the cockpit’s live pulse and signals, work with incidents, manage background jobs and exports, check system health and integrations, and review the platform activity log.

Who This Guide Is For

  • Platform administrators responsible for uptime
  • Operations and on-call staff investigating incidents
  • Engineers triaging failed jobs and exports

Prerequisites

Before you begin, ensure you have:

  • Platform administrator credentials with operations permissions
  • Familiarity with the platform’s normal baseline metrics
  • An incident response process to follow when signals fire

Step 1: Read the Cockpit Pulse

Navigate to Admin → Cockpit.

The cockpit is the platform health intelligence engine. It shows:

  • A live pulse hero with current status
  • Sparklines for errors (24h), failed exports (24h), and dead-letter notifications
  • An integration health grid
  • The signal feed and incident list side by side

Tip: Glance at the three sparklines first — a sudden climb in errors or failed exports is your earliest warning.

Step 2: Review Signals

Navigate to Admin → Cockpit → Signals.

Signals are the raw health events the cockpit ingests. Use this feed to drill into what triggered a change in the pulse before it escalates into an incident.

Step 3: Work with Incidents

Navigate to Admin → Cockpit → Incidents.

Incidents are auto-grouped from related signals. Open an incident to see its timeline and details.

Note: Incident recovery actions on the detail page are part of a phased rollout — some response controls may not yet be wired. Use your runbook to act on incidents while these land.

Step 4: Check Alerts and Failures

Navigate to Admin → Cockpit → Alerts to review active alerting, and use the failures view to see recent failure events the cockpit has captured. The cockpit settings screen lets you tune thresholds.

Step 5: Manage Background Jobs

Navigate to Admin → Ops → Jobs.

This screen lists background jobs and their status. For each job you can:

  • Open it to inspect details
  • Retry a failed job
  • Cancel a running or queued job

You can also clear completed jobs and retry failed ones in bulk.

Tip: Before mass-retrying failed jobs, open one to confirm the root cause is resolved — otherwise they’ll just fail again.

Step 6: Review Exports

Navigate to Admin → Ops → Exports.

Track export jobs and their outcomes here. Failed exports also feed the cockpit’s “failed exports (24h)” sparkline, so reconcile any failures you see flagged there.

Step 7: Check Health and Integrations

Navigate to Admin → Ops → Health to see overall system health and any health incidents. Use Admin → Ops → Integrations to review third-party integration status surfaced in the cockpit’s integration health grid.

Step 8: Review the Activity Log

Navigate to Admin → Activity.

The activity log records platform-wide actions. Use it for forensic review when investigating incidents or support escalations.

Common Mistakes to Avoid

  1. Ignoring rising sparklines: Early signal climbs are cheaper to fix than full incidents
  2. Bulk-retrying without diagnosis: Retrying failed jobs before fixing the cause wastes capacity
  3. Cancelling the wrong job: Confirm the job ID before cancelling a running task
  4. Treating signals as incidents: Signals are raw events; act on grouped incidents with your runbook

Verification Checklist

During an operational review, verify:

  • Cockpit pulse and sparklines are within baseline
  • No unresolved incidents on the timeline
  • Background jobs show no growing failure backlog
  • Exports completed successfully
  • Integration health grid shows all green
  • Activity log reviewed for anomalies

Next Steps