Using CERMM

Alerts and notifications

The server watches every report and raises an alert when something crosses a line. You see them in the bell, and on your own channels.

The rules

An alert clears by itself when the reason goes away, with a margin so it doesn't flap.

Changing the numbers

Settings → Alerts holds the numbers for everyone. People with Manage devices change them. A device can have its own, or turn a rule off: its Alerts section → Alert rules (say, a backup disk that's meant to run full). Changing disk numbers re-judges every disk from its last report at once, and the disk meters colour at the same lines.

Where you see them

• The bell in the title bar (or a): every open alert.
• Red and amber status lights in the Devices list, and Needs attention above it for servers down and critical alerts.
• The disc mark in the title bar: its light is green, amber for an open warning, red for a server down or a critical alert. The browser tab's icon follows.
• A device's Alerts section, with its history.

Notifications

Each person sets up their own channels, in Settings → Alerts, under the rules:

• ntfy: a topic's address, and a token if it needs one.
• Email: through one mail server, set once by someone who manages users, with a test send.
• Webhook: JSON with text and content, so Slack, Discord and Teams can all read it.

Each channel says whether to send critical only or warnings too, whether to send an all-clear, and quiet hours (critical only during them, by the server's clock). Adding one sends a test. Each row shows its last delivery, or the failure in plain words.

How they're sent

Alerts that open, escalate or clear are gathered for 30 seconds and sent as one message per channel, grouped like Needs attention: each server its own line, workstations per problem. Something that opens and clears inside those 30 seconds isn't sent. Suspended people get nothing. Links in a message use the reverse proxy's public address if one is set.