Skip to main content
If you can’t find the solution here, our support team is available at suporte@ainext.com.br
For an endpoint to appear as Online in the portal, several conditions must be met:
  • Authorization: The device must be in Authorized status (Badge “Authorized”).
  • Active Link: The agent maintains a secure WebSocket (WSS) channel.
  • Status Update Logic:
    • AGENT_MAX_OFFLINE_TIME: Default is usually 600s (10 min). If no update is received within this window, the device is marked Offline.
    • TUNNEL_STATUS_CHECK_INTERVAL: The frequency (default 60s) at which the tunnel refreshes connection states.
Rule of Thumb: To avoid “flickering” status, ensure AGENT_MAX_OFFLINE_TIME ≥ (2 × TUNNEL_STATUS_CHECK_INTERVAL) + margin.
Symptoms: UI shows offline even though the agent is installed.Checklist:
  1. Portal Check: Apply filter Status = offline and confirm the device appears. Check the Authorized badge.
  2. Token Validity: If using a Multipurpose Token, ensure it hasn’t expired (tokens are typically valid for 24h).
  3. Network: Verify outbound access to fleet.myrmex.ai, api.myrmex.ai, and myrmex.ai on port 443. Check for TLS inspection/proxies that might break the WSS tunnel.
  4. Local Services:
    • Linux: systemctl status myrmex-endpoint
    • macOS: launchctl list | grep myrmex
    • Windows: Get-Service myrmex_endpoint
Symptoms: Agent version remains unchanged despite portal updates.Causes:
  1. Environment Mismatch: Ensure config.yml points to the correct environment (e.g., prod vs dev).
  2. Artifact Availability: Verify if an update is available for your specific OS/Architecture/Distribution.
  3. Permissions: The supervisor service must run as root (Linux/macOS) or LocalSystem (Windows) to perform the atomic binary swap.
  4. Automatic Rollback: The supervisor automatically rolls back to the previous version if the new one fails health checks. Check logs for “Rollback” entries.
Symptoms: Remote commands (Start/Stop/Scan) time out or fail.Causes:
  1. WSS Instability: If the connection is unstable, the atomic action delivery might fail.
  2. Execution Timeout: High-intensity forensics or deep scans might exceed the default timeout (typically 60-120s).
  3. Authorization: Actions are only executed on Authorized devices.
# View Agent & Supervisor logs
tail -f /var/log/myrmex/myrmex_endpoint.log
tail -f /var/log/myrmex/myrmex_supervisor.log