1.3Core concepts
These terms appear throughout this guide.
| Term | Meaning |
|---|---|
| Monitor | One thing obsrvly watches. It has one type, which you cannot change later. It also has a target, a schedule, regions, assertions and alert channels. |
| Check | One run of a monitor. Each check gives a result: up, down or degraded. |
| Region | Where a check runs from. The dashboard lists the regions you can use, such as us-west and eu-west. The check runs once from each selected region. If you select none, obsrvly uses us-west. |
| Assertion | A rule the reply must meet, such as status_code eq 200. A check is down at the first assertion that fails. See chapter 4. |
| Incident | A record of a problem. By default it opens after 2 failed checks in a row in any one region. It closes when checks pass again in every region that was failing. |
| Alert channel | Where alerts go: email, Slack, Microsoft Teams, Discord, Telegram, PagerDuty or a webhook. See chapter 10. |
| Uptime | The share of checks that passed. The dashboard shows it for 1h, 24h, 7d and 30d. Checks in a maintenance window do not count. |
Most monitors run on their check interval. A heartbeat monitor works the other way round. Your job pings obsrvly, so it has no check interval and no regions. See chapter 5.
If you choose a region that is not available, obsrvly refuses the monitor. Pick a region from the list in the dashboard.
Statuses#
The dashboard shows each monitor with a status. The list on the dashboard page uses short labels. It puts problems first and paused monitors last.
| Status | List label | What it means |
|---|---|---|
| Up | UP | The last check passed. |
| Degraded | DEGR | The check answered, but something is off. For example, it was slower than the monitor's degraded when slower than (ms) limit. Or a certificate or domain expires soon. |
| Down | DOWN | The last check failed, for example no reply, a timeout or a failed assertion. |
| Muted | MUTE | Checks keep running, but the down alert is held while muted. The resolved notice is still sent (3.4). |
| Paused | PAUSE | The monitor does not run. There are no checks and no data. |
| Pending | PEND | The monitor has not run yet. |
| No data | NO DATA | obsrvly has not received a recent result. It does not mean your site is down. |
| Unknown | UNKN | The dashboard could not load the results right now. |
The status mark also has a shape. Up is round. Degraded is a rotated square. Down is a square. No data is a dashed circle.
Incidents#
Open alerting → incidents to see every incident. Use the filters all, firing and resolved. An open incident shows firing and not yet resolved.
- Select Acknowledge to tell your team you are working on it. Reminders for that incident stop. The card then shows
acknowledged by <your email>. - Acknowledgement notices go to Slack, Microsoft Teams, Discord and Telegram channels only. In PagerDuty, the incident is acknowledged. Email and webhook channels do not get them.
- When the monitor passes again, obsrvly still sends the resolved notice.
An empty incidents page is good news. It means nothing has failed enough to open an incident.
Mute or pause a monitor#
Open a monitor to find these buttons at the top:
- Mute keeps the checks running, but the down alert is held while muted. The resolved notice is still sent (3.4). Choose Mute for 1 hour, Mute for 4 hours, Mute for 24 hours or Mute indefinitely. Select Unmute to end it early.
- Pause stops the monitor completely. Select Resume to start it again.
Chapter 3 covers pausing, muting and maintenance windows.