> ## Documentation Index
> Fetch the complete documentation index at: https://docs.marcus.involvecloud.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Handling alerts

> What raises an alert, and how to acknowledge and resolve one.

An alert is raised when something needs attention, most often a device or Collector that has stopped responding, or a room that failed its readiness check. The **Alerts** item in the sidebar shows how many are open, and turns red when any is critical.

## What raises an alert

| Alert | Raised when | Clears |
| - | - | - |
| **Device offline** | A device stops responding. Warning by default | Automatically, when it responds again |
| **Collector offline** (`collector_offline`) | A Collector stops checking in. Critical, because none of its devices can be monitored; Warning when a standby Collector has taken over its devices, or when the offline machine is the standby itself | Automatically, when it checks in again |
| **Readiness check failed** (`readiness_failed`) | A room's [readiness check](https://docs.marcus.involvecloud.com/portal/room-readiness.md) fails. One alert per room, naming the failed steps. Warning | Automatically, when the room's next check passes |
| **Power cycle failed** (`power_cycle_failed`) | A room's overnight power cycle fails, for example a device doesn't come back on. Warning | Automatically, when the room's next power cycle succeeds |
| Device-specific alerts | A device reports a problem, depending on the device | When resolved |

On the **Alerts** page each alert shows its key in small capitals, such as `readiness_failed`, with the message beside it. Alerts for a room link to the room, and device alerts to the device.

Alerts are also closed automatically if their device is deleted.

Devices that [Room Readiness](https://docs.marcus.involvecloud.com/portal/room-readiness.md) has switched off for the night don't raise offline alerts until the morning warm-up is over.

## Severity

| Severity | Shown in | Use |
| - | - | - |
| **Critical** | Red | Needs action now: a room or site is affected. |
| **Warning** | Amber | Needs attention soon. |
| **Info** | Blue | For awareness. |

## Work through alerts

**Alerts** has four tabs: **Open** (the default), **Acknowledged**, **Resolved** and **All**. Each alert shows the device (or the room, for a readiness or power cycle alert), what happened, when it opened and the details.

* Select **Ack** to show you're dealing with it. The alert moves to **Acknowledged**, recording your name and the time, so colleagues don't duplicate the work.
* Select **Resolve** when it's fixed and won't clear on its own.

Offline alerts don't need resolving by hand: they clear the moment the device or Collector is back.

You need the **Acknowledge alerts** or **Resolve alerts** permission for these actions.

## Get alerts outside the portal

Alerts can also be sent by email, to a Microsoft Teams channel, to any system that accepts a webhook, or straight into ServiceNow as incidents. An administrator sets these up; see [Notifications](https://docs.marcus.involvecloud.com/admin/notifications.md).

Notifications go out when an alert first opens, not every time it repeats, so an unstable device doesn't flood your inbox.

## How long alerts are kept

Resolved alerts are kept for 12 months after they're resolved, then deleted. Open and acknowledged alerts are kept until they're resolved.

Updated 06 Nov 2026 14:07 · page 22d1079


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.