Alert Rules

This page explains when go-web-monitor sends an offline alert, how the failure count accumulates and resets, and what the alert email contains.

Failure Count

Each site has a count field in .webMonitor.json that records consecutive offline checks. After each round it updates as follows (internal/service/monitor.go:159):

Result this round	Previous count	Action	New count
Online	Any	None	0
Offline	0–3	None	Previous + 1
Offline	4	Send alert	0

The condition is written as (count+1)%5 == 0; because the count resets on the 5th failure, the effective rule is "one email per 5 consecutive offline checks".

stateDiagram-v2
    [*] --> Healthy
    Healthy: Online (count = 0)
    Failing: Offline (count 1–4)
    Healthy --> Failing: Offline
    Failing --> Failing: Offline and count < 4
    Failing --> Healthy: Online
    Failing --> Alerted: 5th offline
    Alerted: Alert sent, count reset
    Alerted --> Failing: Still offline
    Alerted --> Healthy: Online

Alert Frequency

Situation	Result
Site goes offline for 1–4 minutes, then recovers	No email
Site stays offline	One email at roughly minute 5, 10, 15, and so on
Manual `refresh`	Counts as a round and speeds up accumulation

No "recovered" email is sent when a site comes back online.

Alert Email

Field	Value
Subject	`Website Offline Alert: <url>`
Format	HTML (`Content-Type: text/html; charset=UTF-8`)
Body	URL, `Status: Offline`, `Reason: HTTP <code>`, check time

On a connection failure the status code in Reason is HTTP 0.

Delivery Conditions

Alerts call SMTP.SendEmail() in a background goroutine. When any of the following is false, no email goes out and the error is written only to stderr (slog):

enabled is true
host is set
to has at least one recipient

See Email Alert Setup for configuration and SMTP Connection for connection details.