Router health and diagnostics

Diagnose a router before changing it

Open the router detail page and read several signals together. Centipid Billing can show reachability history/incidents, live bandwidth/system metrics, top consumers, service state, monitoring, trends, collections, and diagnostic findings.

Signal interpretation

Signal	What it tells you	What it does not prove alone
Online/reachability	The router’s management tunnel (SSTP) is up and it answers	That every subscriber can authenticate, or that local LAN/Hotspot ports are correct
Active sessions	RADIUS accounting sessions	That all sessions are current or healthy
CPU/memory/system metrics	Router resource pressure	The cause of a payment/access failure
Bandwidth/top consumers	Current traffic distribution	Package-policy correctness without subscriber context
Monitoring findings	Checks detected by the diagnostic pipeline	That every finding is safe to auto-remediate
Daily collections	Payments associated with the router/context	Provider settlement or bank reconciliation

Run diagnostics

1. Verify the router identity and incident time.
2. Choose Diagnose.
3. Wait for findings to complete.
4. Review Issues, Passed, and the full result set.
5. Open a finding and read its evidence and proposed remediation.
6. Apply remediation only when it matches the observed failure.
7. Re-run the check and an end-to-end subscriber test.

Targeted recovery actions

• Sync router time: correct clock drift that affects certificates, schedules, accounting, or logs.
• Sync walled garden: refresh captive destinations needed for payment/login flows.
• Redownload Hotspot: refresh portal assets/configuration when the router copy is stale.
• Reprovision: give the router a new tunnel identity and a fresh script to paste. Use it when the management tunnel will not come back or setup is incomplete; the router stays offline until the new script runs.
• Reboot: use only after understanding service impact and having a recovery path.

Monitoring and alerts

Add the router to monitoring, then configure Settings → Operator alerts. Verify recipient numbers and whether WhatsApp is preferred. Alert delivery depends on communication-provider readiness; a monitoring finding and a delivered SMS are separate states.

High CPU or bandwidth

Use live metrics and top consumers to identify timing and scope. Check whether load corresponds to expected traffic, a package without limits, a subscriber/device anomaly, or router process pressure. Avoid changing every package or rebooting before capturing the active evidence.

Escalation packet

Provide support with sanitized evidence:

• the ISP and router identity (not secrets);
• start/end time and timezone;
• affected access type and approximate subscriber count;
• reachability, management tunnel, and service state;
• diagnostic finding text;
• RouterOS version where available;
• one test subscriber result;
• actions already attempted and their outcomes.