# How to pause downtime monitoring while you work on a host

When you deploy an update or do maintenance that takes a site down for a while, the monitoring
reports that downtime. The Hosted Power support team then follows it up. **Pause Downtime
Monitoring** in the host header stops that for a period you choose. Afterwards the checks switch
back on automatically.

## Before you start

- **Access to the host** in the TurboStack Platform. Anyone who can open the host can pause its
  monitoring.
- **An idea of how long the work takes.** You choose 1, 2, 4 or 8 hours, or 1 day. Keep it as
  short as you can: real problems during the pause are not reported either.

## Pause the monitoring

1. Open the host and select the bell icon in the header, between the Revisions (clock) icon and
   **Save**.
2. Read the summary in the popup. All checks of this server are paused: websites, services, disk,
   load and more.
3. Choose a period under **Pause for**.
4. Select **Pause monitoring**. The popup closes after a few seconds.

![The Pause Downtime Monitoring popup with the period selector|1000](../../assets/screenshots/platform/hosts/pause-monitoring_popup.png)

The pause needs no publish and changes nothing in the host's configuration.

## While the pause runs

- The bell in the host header turns orange. Hover it to see until when the monitoring is paused.
- The Hosts list shows an orange **Monitoring paused** badge next to the host name.
- Selecting the bell again shows who paused the monitoring and until when.

![The popup during a pause, with the end time and the Resume monitoring button|1000](../../assets/screenshots/platform/hosts/pause-monitoring_paused.png)

![The Hosts list with the Monitoring paused badge on a host|1000](../../assets/screenshots/platform/hosts/managing-hosts_monitoring-paused.png)

## Resume the monitoring

The monitoring resumes by itself when the period ends. When you finish earlier, select the bell
and then **Resume monitoring**, so real problems are reported again. Anyone with access to the
host can resume it, not only the person who paused it.

## Verify

Open the Revisions (clock) icon and the **Actions** tab of the [History](revisions.md) view. Every
pause is listed there with who started it, the duration, the planned end and a status: **Paused
now**, **Resumed early** or **Ended on time**.

![The Actions tab of the History view listing a running pause|1000](../../assets/screenshots/platform/hosts/revisions_actions.png)

## If it did not work

- **"This server is not known in our monitoring, so there is nothing to pause."** - the host
  is not monitored yet, for example right after it was created. Nothing was paused.
- **"Monitoring could not be updated"** - nothing was changed. Try again in a minute, or
  [contact support](../support.md).
- **The bell stays orange after Resume monitoring** - the resume failed and the pause is still
  active. Try again. If it keeps failing, [contact support](../support.md) so the alerts are
  switched back on.

## Related

- [History - revisions, deploys, cloning and actions](revisions.md) - the Actions tab
- [Hosts](index.md) - the host header and the Hosts list
- [Health (per host)](health.md) - the checks that are paused
- [Publishing changes](publishing.md)