Responding to alerts
When a device crosses a threshold, goes quiet, or otherwise needs attention, the platform raises an alert. You'll see alerts in two places: the notification bell at the top of the app, and (if your provider has set it up) in your email inbox.
This page covers what alerts look like, what the colors mean, and how to acknowledge them.
The notification bell
The bell icon sits in the top-right of every page. A red badge on the bell shows the number of unread notifications, up to 99+.
Click the bell to open a dropdown with your most recent unread notifications. Each one shows:
- A severity tag (Info, Warning, or Critical), color-coded.
- A title (typically the name of the alert and the device).
- A short message, if the alert included one.
- When it arrived, as a relative time ("5 minutes ago").
Click See all notifications at the bottom of the dropdown to open the full inbox.
The notifications inbox
The inbox page shows every notification the platform has raised for you, grouped by Today, Yesterday, This week, and Earlier.
You can narrow what you're looking at with the filters at the top:
- Read status: unread, read, or both.
- Severity: any combination of Info, Warning, and Critical.
- Date range: a custom window.
The inbox loads up to 200 notifications at a time. If you need to look further back, narrow the date range.
What the severities mean
The three severities are set by whoever built the rule that raised the alert. As a rough guide:
| Severity | What it usually means |
|---|---|
| Info | Worth noting, no action required. ("Device came back online.") |
| Warning | Worth attention. Something is drifting toward a limit. |
| Critical | Act now. A threshold has been breached or a device has gone silent. |
If a particular alert seems to use the wrong severity, talk to your provider; they can adjust the rule. If your plan includes workflow editing, you may be able to adjust it yourself, see Creating workflows.
Acknowledging an alert
Acknowledging tells the platform, and everyone else watching, "I've seen this and I'm on it." It clears the device's alert state for every viewer, your name is recorded (the Status widget briefly shows "Acknowledged by ..." so a colleague opening the dashboard knows who took it), and the alarm's notifications pause: while the condition is still being crossed, the workflow stops emailing, texting, and adding inbox rows for that device. Device actions such as downlinks keep running. The moment the condition actually clears, the alarm recovers; if it later trips again, that's a new alarm and notifications resume in full.
While a pause is active, the Status widget says so: "Notifications paused: acknowledged by ..." with a Resume notifications button, so an acknowledge is never invisible and can always be undone. Resuming does not re-raise the alert state by itself; the workflow's next fire does that normally. A pause also ends on its own if the workflow's condition stays quiet on that device for 24 hours: a long-silent alarm is treated as over, and the next trip opens a fresh, fully-notified one. Readings that keep the condition active count as activity even when a schedule window or throttle is holding the actions back, so a business-hours workflow's acknowledgment survives the weekend.
You acknowledge from a dashboard: click Acknowledge on a Status widget that's showing the alert. Any device dashboard or shared dashboard with a Status widget works.
Reading a notification in the inbox is separate: Mark as read only clears it from your unread badge, it does not acknowledge the device's alarm. The two used to be linked; they no longer are, so ticking through your inbox can't silently clear an alarm your team still needs to act on.
Permission required: any signed-in user can read the inbox. Acknowledging requires Operator or Workspace Admin (read-only users see the alert without the button).
Acknowledging quiets the alarm but does not fix the underlying condition - and the quiet lasts only until the alarm recovers and trips again. To stop a persistently noisy workflow for good, adjust its conditions or add a Throttle node (see Creating workflows), or if it's a provider-managed workflow, mute it for your account from the Workflows list.
Alert emails
If you've been added to a rule's notification group, you'll receive an email when that rule fires. The email contains:
- The severity, rule name, and device name in the subject.
- A short summary in the body, with the readings that triggered the alert and the thresholds they crossed.
- The time the alert fired, shown in your own timezone and time format as set in Account settings. If you haven't set a timezone, the time is shown in UTC and labeled as such.
- A button to Open the rule in the platform.
The email is sent from and branded by your provider. There is no acknowledge button in the email itself; open the platform to acknowledge.
If you no longer want to receive emails for a particular rule, ask a Workspace Admin to take you out of the notification group it uses; they manage groups under Administration -> Notification Groups (see Notification groups). If the rule is one your provider manages for you, its recipients belong to the provider, so ask them instead. To stop all alert emails to yourself at once, switch off Email notifications in your account settings; there is no per-rule unsubscribe link in the email itself.
Alert states on a device
A device can be in one of three alert states at a time: Info, Warning, or Alert (also shown as "Critical"). When no rule has flagged the device, it has no alert state and shows as normal.
You'll see the current alert state:
- As a colored pill on the device list, next to the device name. See Finding your devices.
- On any Status widget on a dashboard that's been pointed at that device.
- On the device detail page.
Alert states clear when:
- An Operator or Workspace Admin acknowledges them.
- The rule that set the alert has an expiry, and that time has passed.
- The rule itself is changed or removed by your provider.
Snoozing and muting
You can't snooze or mute individual alerts from the inbox. If a particular rule is too noisy or too quiet, your provider can adjust its thresholds, severity, or schedule. If your plan includes workflow editing, you can adjust a workflow you own yourself, see Creating workflows.
Frequently asked
I marked a notification as read by mistake. Can I undo it? You can still find it in the inbox by switching the read-status filter to "Read" or "All". You can't un-mark it as read, but the original details remain.
I see an alert in my email but not in the bell. Why? Email goes to the members of the rule's notification group, which can include people outside the platform, while the bell shows notifications raised for your own account. Being in one doesn't put you in the other. A Workspace Admin can check the group's members; if the rule is managed by your provider, ask them.
The bell shows 0 but a device on my dashboard is red. Why? The dashboard pulls live state directly from the device. The bell only shows notifications that were raised for your account specifically. Ask a Workspace Admin (or your provider, for a provider-managed rule) which notification group should include you.