Email alerts
Who receives an email alert, what the message contains and how to read the notification log when one doesn't arrive.
Last reviewed
On this page
Email is the channel every site starts with. This page describes who receives it, what a message contains and how to read the log when one doesn't arrive. The rules that decide when a message is sent are on the alert rules page.
Set up an email notification
Open the site's notifications
Open the site and choose Notifications. Choose Add Notification, or Add First Notification on a site with none.
Name it and keep Email as the type
Enter a Notification Name, for example Team Email Alert. Notification Type starts as Email.
Add recipients
Enter one email address per row. Any valid address works, including people who don't have an OctoPing account. The form won't save while an address is malformed, and a notification needs at least one address.
Choose the triggers and delivery settings
The four switches and the Retry Threshold and Notification Interval start on the defaults described on the alert rules page.
Create it, then choose Send Test
The test goes out immediately and the result appears at the top of the page. It's also written to the log with the trigger test.
A heartbeat has its own Notifications page with the same form and two switches. A new site or heartbeat already has a Default Email Notification that emails the workspace's members as they were when it was created.
What arrives
- The sender name is OctoPing Alerts. Account messages such as invitations and password resets come from OctoPing, so a filter on the sender name separates alerts from them.
- A notification with several recipients sends one message addressed to all of them.
- Down alerts, recovery alerts, content change alerts and robots.txt change alerts are sent as an HTML message with a plain-text version. Every other alert, and every heartbeat alert, is plain text.
- A long error message is cut at 300 characters in the body.
The subject line
Most subjects start with an emoji that shows the kind of alert, then name the site, shortened to its first 50 characters (40 for a DNS change). The checker blocked notice starts with Notice instead, and a heartbeat alert names the heartbeat. A down alert says the site is down. When only some of the checking locations see the failure, it names them, and it ends with the failure reason from the failure reasons list. A recovery alert says the site is back online and, when OctoPing knows the length of the outage, how long it lasted.
What a down alert contains
The plain-text version of a down alert lists these parts.
| Part | What it says |
|---|---|
| Problem | The failure reason, with its plain-language explanation. |
| HTTP status | The status code and its name, when the failing check received a response. |
| Response time | The failing check's response time, when there is one. |
| Error | The error text from the failing check, when there is one. |
| Consecutive failures | How many checks in a row have failed. |
| Last checked | The time of the latest check. |
| Checked from | Every checking location with its latest result: up, down or no recent data, with the status code and how long ago it ran. |
| A link | A link to the site in OctoPing, opened on its downtime view. |
A recovery alert has the same shape. It reports the length of the outage, when it started and the cause, and lists the current result from each location.
When a message doesn't arrive
| In the log | What to check |
|---|---|
| Sent | OctoPing handed the message to its mail provider without an error. Look in the recipient's spam folder, then in their mail server's own logs. A message that the receiving server rejects later doesn't appear in the log. |
| Failed with No valid email recipients found | The notification has no addresses left. Edit it and add one. |
| Failed with another message | The mail provider refused the message. The Error column quotes what it said. |
| No row at all | No alert was due. The notification may be disabled, the matching switch may be off, the failure count may be under the Retry Threshold or the interval may not have passed. A held-back alert inside a maintenance window appears as Skipped. |
Send Test tells a delivery problem apart from a rule that didn't fire. A test that arrives means the address and the channel work, and the question is about the rule.
Related
- DocsAlert rulesHow a failed check becomes an alert: the retry threshold, the repeat interval, the four triggers, recovery, the default notification, the log and what a maintenance window holds back.
- DocsSlack alertsConnect Slack to a workspace, post alerts to a channel, and what the log shows when a message fails.
- DocsAdd your first site, alert and status pageAdd a site, get a first alert through the default email notification and publish a status page, using the app's own labels.
- FeatureWebsite alerts by email, Slack, Discord, webhook and pushChoose where alerts go, how many failures it takes to send one and how often reminders repeat. Recovery messages and a log of every attempt are included.