Skip to main content

FAQ and Common Issues

Quick diagnosis and resolution for common user-facing problems.

Before you troubleshoot​

Check these first:

  1. Are you logged in with the correct role?
  2. Is the API reachable at http://localhost:8000?
  3. Is the dashboard reachable at http://localhost:3000?
  4. Are required backend services healthy?

Login and access issues​

Login fails with unauthorised error​

Possible causes: wrong credentials, expired token or session, misconfigured environment values.

  1. Retry login and verify username and password.
  2. Log out and log in again to refresh the token.
  3. Ask an admin to verify your role and account status.

Cannot see pages other users can see​

Possible cause: role-based permissions differ. Check Role-Based Access and request the required role from an admin.

Data and refresh issues​

Dashboard cards do not update​

Possible causes: API call failure, WebSocket disconnected, upstream service lag.

  1. Refresh the browser once.
  2. Confirm the API health endpoint responds.
  3. Confirm the alerts stream is receiving events.

Alerts appear duplicated​

Possible cause: source systems sending near-identical events across channels. Check dedup hash fields in alert detail. Group by source and timestamp to verify true duplicates.

Investigation and response issues​

Investigation stays pending too long​

Possible causes: worker queue backlog, agent endpoint timeout, dependency degradation.

  1. Check worker and agent service health.
  2. Retry the investigation once backlog clears.
  3. Escalate to platform engineer if persistent.

Open the Agent Monitor (Agent Fleet page) to confirm the fleet is up. If the agent and MCP tool servers are not running, every agent shows "Unreachable" and the MCP domains show "down", as below. Investigations will sit pending until those services are back. Start the agent and MCP servers (or their containers), then retry.

Agent Monitor with agents unreachable and MCP domains down

Playbook execution blocked​

Possible causes: missing permission, approval required, playbook disabled.

  1. Verify your permissions and role.
  2. Check the approvals queue and request status.
  3. Confirm the selected playbook is enabled.

CPS and IoT issues​

Device shows untrusted attestation​

Open device details, compare firmware version and hash with expected baseline, escalate and isolate if policy requires.

Physical-cyber correlation seems missing​

Check the correlation view and time window. Confirm both cyber and physical signals are present for the same period.

Frequently asked questions​

Is AuroraSOC safe to use in production as-is?​

AuroraSOC includes production-oriented controls. Review deployment hardening guidance before full production rollout.

Can I run only part of the platform?​

Yes. Run core API and dashboard flows first, then progressively enable agents and advanced integrations.

Where do I find engineering-level diagnostics?​

See the developer documentation sections on monitoring and observability.

Escalation matrix​

SeverityExample ConditionEscalate ToTarget Response
LowUI rendering issue without data impactSOC operator leadSame business day
MediumDelayed investigation resultsPlatform engineer on-callWithin 2 hours
HighPlaybook execution blocked during active caseSecurity engineering leadWithin 30 minutes
CriticalWidespread auth failures or API outageIncident commander / SREImmediate

Data to capture before escalation​

  1. Timestamp and timezone of issue occurrence
  2. User role and affected page or API endpoint
  3. Request ID or trace ID if available
  4. Screenshot, error message, and reproduction steps
  5. Whether the issue is reproducible or intermittent