Manage devices

Work with devices in NervesHub: tags, the device detail tabs, and actions like reboot, reconnect, identify, and clearing the penalty box.

Written By Josh Kalderimis

Last updated About 2 months ago

A device is a single unit connected to NervesHub, identified by a unique identifier (usually a serial number). This guide covers day-to-day device management in the web UI.

For getting a device connected in the first place, see the Quickstart and Production setup with NervesKey.

The device list

A product's Devices page lists every device with its status, tags, current firmware, and last-seen time. Filter and search to find devices by identifier, tag, firmware version, or connection status.

You can create devices from the UI or the CLI, individually or in bulk:

nh device create
nh device bulk_create

Tags

Tags are free-form labels on a device. They're how deployment groups target devices, so a consistent tag scheme (for example field-test, eu, cellular) is worth planning early.

Set tags in the device's settings, or from the CLI:

nh device update <identifier> tags production eu

Device detail tabs

Opening a device shows several tabs:

Tab

What's there

Details

Identifier, status, tags, current firmware, metadata.

Health

Reported metrics, alarms, and network interface data.

Firmware history

Which firmware the device has run over time.

Logs

Device logs (requires the ClickHouse-backed logging).

Activity

Audit log of actions on this device.

Console

Live remote IEx session (if enabled).

Local shell

Live system shell (if enabled).

Settings

Tags, description, deployment group, and device actions.

See Device health, metrics & geo and Remote console & local shell for those tabs.

Actions

From a device's page you can:

Action

What it does

Reboot

Ask the device to reboot.

Reconnect

Force the device to drop and re-establish its connection.

Identify

Trigger a physical identify signal (for example an LED), if the firmware supports it.

Clear penalty box

Release a device that was penalized after repeated update failures, so it can update again.

Pin / unpin

Keep a device handy at the top of your view.

Move to deployment group

Assign the device to a specific deployment group.

Toggle firmware updates

Pause or resume updates for this device from its deployment.

Many of these are also available from the CLI (nh device ...) and the HTTP API.

Penalty box

When a device fails to update too many times, the orchestrator puts it in the penalty box for a configured timeout, so it stops retrying aggressively. It rejoins updates when the timeout passes, or immediately if you clear the penalty box. The thresholds and timeout are set per deployment group. See Deployment orchestration.

Deleting and restoring

Deleting a device is reversible for a period: a soft-deleted device can be restored. While soft-deleted, its certificate validation is skipped and it can't authenticate until restored (or fully destroyed). This is why a deleted device won't reconnect until you restore it.

Related