Guided pilot: How to Suspend Internet Automatically When Invoices Are Unpaid
An operator guide to turning overdue invoices into controlled access decisions, with grace periods, payment checks, and proof of reconnection.
Run the workflow with proof
Start with a billing decision, then change access
Evaluating automatic suspension in a guided pilot needs more than a nightly list of overdue invoices. Dunning is the sequence of reminders and follow-up after a payment becomes due. An access policy translates the current billing evidence into a service decision: keep access active, allow grace, limit service, suspend, or retain a manual hold.
Keep the invoice, customer, service, and network identity separate. One customer may have several services; an overdue installation charge should not silently suspend every connection. Decide which charges affect which service before enabling a scheduler.
Architecture: dunning to enforcement
Invoice + payment allocations + manual holds
|
Dunning and grace rules
|
Service-specific policy preview
|
Approval + fresh balance check
|
Queued access job
|
Qualified enforcement adapter
/ \
RADIUS policy / session MikroTik service rule
\ /
Result + access check
|
Support evidence trail
Treat policy intent, adapter completion, and observed access as separate facts. A job marked complete does not prove that a subscriber has reauthenticated or can pass traffic.
RADIUS authenticates a session and supplies authorization attributes. Changing future authorization does not necessarily change an existing session. Where the NAS supports it, CoA can request a session change and Disconnect can terminate a session so the subscriber authenticates again. Match the NAS and session identifiers, and record ACK, NAK, or timeout. These mechanisms are defined in RADIUS Dynamic Authorization, RFC 5176.
A MikroTik deployment may instead enforce a service rule through a queue, profile, or address list. Qualify the exact rule and restoration behavior in a lab. Do not assume deleting a session or changing a generic firewall rule is a reversible suspension workflow.
Workflow example: an overdue service is paid during grace
This is an illustrative runbook, not a customer result or an ISPAgents automation guarantee.
- An invoice for service
LAB-SVC-Abecomes overdue. Dunning sends the configured reminder, and the policy enters a 48-hour grace period. The grace period is an example operator choice. - The preview records the outstanding amount, due time and timezone, service identifier, enforcement target, and any dispute or manual hold.
- A verified payment arrives before grace expires. Allocate it to the relevant invoice, then recompute the policy. Cancel or supersede any stale suspension intent.
- If the balance remains unpaid when grace ends, recheck it immediately before executing an approved suspension job. Record the previous access state and the intended restriction.
- When a later payment clears the required balance, queue restoration only if no independent hold remains. Restore the recorded service profile, then verify the session and expected access.
- If payment is recorded but the subscriber remains offline, open an exception for support. Do not mark reconnection successful solely because the adapter returned success.
For partial payments, define the eligibility threshold explicitly. For reversals, recompute eligibility through the same policy path rather than issuing an immediate blind disconnect.
Operational risks
| Failure | Control to establish before automation |
|---|---|
| Payment arrives after a batch preview | Revalidate balance and policy version at execution; discard stale jobs. |
| Retry repeats a destructive action | Use a stable job identity and compare desired state with observed state. |
| Wrong customer or NAS is selected | Require a verified service-to-network mapping; stop on ambiguity. |
| Router is unreachable | Keep the result pending or failed, alert an operator, and reconcile before retry. |
| A manual hold is overwritten | Give holds explicit precedence and require an authorized release. |
| Restoration reports success but traffic fails | Check session and service evidence; route paid-but-offline cases to support. |
Maintain an emergency bypass procedure owned by an authorized operator. Record its scope and expiry so a temporary exception does not become permanent access drift.
Implementation checklist
- Define due dates, timezones, reminders, grace, partial-payment rules, and dispute handling.
- Map every eligible service to its specific network identity and enforcement target.
- Run previews against mock invoices, including payment during grace and payment during execution.
- Validate RADIUS or MikroTik behavior for the actual NAS, firmware, and access method.
- Exercise duplicate jobs, timeouts, stale previews, and unreachable devices in the lab.
- Record previous state, approval, execution evidence, and a tested rollback procedure.
- Start with an operator-approved canary and review restoration evidence before expanding.
Where ISPAgents fits
ISPAgents offers service-policy preview, queued jobs, approvals, no-op handling, result recording, and a mock-safe worker as a guided pilot. Live enforcement remains gated on real adapters, credential resolution, rollback evidence, and monitoring. Subscriber billing and payment allocation are also guided-pilot foundations, not a complete accounting suite.
The existing FreeRADIUS path is a guided pilot with a signed Status-Server probe. Opt-in SQL accounting import has accepted proof. CoA/Disconnect drivers have fake-transport test coverage; real NAS mutation and end-to-end policy enforcement remain unqualified. SQL/API sync and policy write-back are not a complete integration. MikroTik RouterOS REST operations are a guided pilot; guarded writes and packaging remain product work. Qualify an enforcement path before promising automatic reconnection.
Related solutions and guides
Read the automatic internet suspension scope, MikroTik ISP billing requirements, and FreeRADIUS versus managed RADIUS pilot guide. For the payment evidence that feeds the policy, use the mobile-money reconciliation guide for pilot evaluation.
Continue the operations map.
One Subscriber, One Device Identity
Available: The canonical identity foundation supports mapping protocol observations toward one device/customer path using strong serial/MAC evidence. It does not establish universal billing, RADIUS, or controller integration.
Open pageGuidesFair-usage policies for unlimited ISP plans
Keep access available after a usage threshold, explain the reduced speed before purchase, and give support a policy they can defend.
Open pageSolutionsAutomatic Internet Suspension Software — Guided pilot
Guided pilot: Preview, queued jobs, approvals, result recording, and a mock-safe worker can be evaluated in a guided pilot. Live router/RADIUS mutation remains gated; automated suspension and restoration are not a shipped promise.
Open pageSolutionsCustomer Self-Service Foundations — Guided pilot
Guided pilot: Field-job lifecycle, serialized stock, basic tickets and subscriber portal flows have narrow synthetic-data proof and remain guided pilots. Physical installs and invite email delivery are unproven. The technician PWA, dispatch board, dedicated stock workspace and IPAM remain roadmap.
Open page