Skip to main content

Service Availability Monitoring

Monitoring lets you automatically check the availability of external services (HTTP or PING) and immediately notify responsible team members via a voice call when outages occur. The system works together with billing and telephony: when an issue is detected, 4phone automatically calls the designated contact and asks them to confirm that the incident has been acknowledged.

How monitoring works​

The platform periodically checks the specified URL or host. If several consecutive checks return an error, an incident is created and the escalation chain is triggered: calling the list of responsible contacts with a voice alert.

Core entities:

  1. Contact points define who to call and in which language.
  2. Monitors define what to check, how frequently, and how to respond.
  3. Incidents are created automatically after a series of failed checks.

Contact points​

A contact point links a specific person to a notification channel. Before creating a monitor, you must set up at least one contact point.

Fields when creating a contact point:

  1. Name to identify the contact in the list.
  2. Type: internal voice (call to an extension) or external voice (call to a mobile or landline number).
  3. Notification language: Russian or Uzbek. The voice message will be generated in the selected language.
  4. Number: for the internal type, select an extension from the list; for the external type, enter a phone number.
  5. Active status: inactive contacts will not be used during escalation.
Important

For internal calls, the operator must be online in the softphone. For external calls, the tenant must have an active outbound SIP trunk.

Creating a monitor​

A monitor defines the check target and the rules for responding to failures.

Basic parameters​

  1. Name to identify the monitor in the list.
  2. Check type: HTTP (checking a URL and expecting a specific response status code) or PING (ICMP host check).
  3. Target: URL for HTTP (for example, https://example.com/health) or hostname/IP for PING.
  4. Check interval: how often to run the check (minimum 30 seconds).
  5. Timeout: maximum time to wait for a response.
  6. Expected status code (HTTP only): which HTTP status code is considered successful (200 by default).

Thresholds and behavior​

  1. Failure threshold: how many consecutive failed checks are required to open an incident.
  2. Recovery threshold: how many consecutive successful checks are required to resolve an incident.
  3. Cooldown: minimum time (in seconds) between repeated notifications for the same monitor.

SIP trunk assignment​

For voice notifications, the monitor requires an assigned SIP trunk. Outbound escalation calls will be placed through this trunk.

Alert messages​

Separate alert texts can be configured for Russian and Uzbek. These messages will be spoken during the voice call. If no custom text is provided, the system uses a default template with the monitor name.

Escalation chain​

Defines the order and recipients for incident calls:

  1. Contact: selected from your configured contact points.
  2. Order: determines the sequence of calls.
  3. Delay: how many seconds to wait before calling this step.
  4. Retry count: how many times to retry the call if the contact does not answer.

Statuses​

Monitor​

  • Healthy: the last check succeeded.
  • Failed: the last check returned an error.

Incident​

  • Open: a failure was detected and notifications are being sent.
  • Acknowledged: the responsible person answered the call and pressed key 1 to confirm.
  • Resolved: the service recovered (successful checks reached the recovery threshold).

How voice notifications work​

When placing a call, the system:

  1. dials the contact's number;
  2. plays the generated voice message;
  3. asks the recipient to press 1 to acknowledge the incident;
  4. if the recipient presses 1, the incident transitions to the "Acknowledged" status;
  5. if the recipient does not answer or does not press 1, the system proceeds to the next step in the escalation chain.
Check

Voice notifications are billed. Each call is recorded in billing as a separate monitoring operation.

What to verify​

Before launching monitoring
  1. at least one active contact point is created;
  2. the tenant has an active SIP trunk for outbound calls;
  3. the trunk is assigned to the monitor;
  4. the escalation chain contains at least one step;
  5. the test URL has been verified and incidents are created properly;
  6. the recipient knows to press 1 to acknowledge the incident.

Common issues​

  1. Call is not received: verify that the SIP trunk is active and assigned to the monitor, and that the contact is active and has the correct phone number.
  2. Incident is not created: make sure the failure threshold is greater than 1 and the target service is actually unreachable.
  3. Status does not change to "Acknowledged": the recipient must press key 1 during the call.
  4. "Failed to connect": check the URL/host, network reachability, and DNS resolution.

Next, we recommend reading How to Connect a SIP Trunk and Common Issues and Quick Fixes.