Skip to content

Alerts

Alerts tell the right people when something goes wrong, so you hear about a crashed server before your players do. Admins decide what alerts and who hears about it; each person decides how they’re told.

The bell at the top right of every page is your inbox: every alert sent to you, from every org you’re in, for 30 days, each linking to the page it’s about. The count on it is what’s new since you last opened it.

Open the bell, then its settings, to choose where else your alerts go:

  • Email, to your account’s address (on from the start; turn it off if you like).
  • Telegram, ntfy (push to your phone), Discord, Slack, or a webhook.

Each has a Test button. Only you see your channels. One that fails for 3 days is turned off, and you’re emailed.

On Settings → Alerts (admins and owners), each rule is:

  1. A condition (below),
  2. where it applies: every node, one node, or one server,
  3. who it tells: people by role (owners, admins, members) or by name, and
  4. team channels it also posts to, if any.

Members are only told about servers they have access to, and never about database instances: those alerts go to owners and admins, whoever else a rule names.

New orgs start with rules that tell their owners about crash loops, offline nodes, and low disk.

Server Node
Crashed Offline (after 5 minutes, or the grace you set)
Down after repeated crashes Disk running low
Install finished or failed Wings updated or rolled back
Backup failed Signed actions and passkey changes
Restore finished or failed CPU, memory, or disk above a percent, for a while
Over its disk limit
CPU, memory, or disk above a percent, for a while

Database instances (on the node’s Databases tab):

  • Down or restarting: it stopped, or kept restarting. Raptor starts it again by itself, and tells you once it answers again.
  • Disk above a percent of its disk limit, for a while. Instances without a disk limit aren’t checked.
  • Connections above a percent of its max connections, for a while.
  • A version change, patch update, or going back finished or failed.
  • A backup of the instance failed (Back up an instance).

They need Wings v0.11.0 or later on the node.

Conditions that end (a node back online, a disk with room again, an instance back up) say so too, unless the rule says not to.

Places the whole team reads, added on Settings → Alerts: Discord and Slack (webhook URLs), Telegram (a group with Raptor’s bot), ntfy (a topic), and webhooks.

Webhooks are JSON, signed: X-Raptor-Signature: sha256=<hex> is the HMAC-SHA256 of <timestamp>.<body> with your secret, and X-Raptor-Timestamp and X-Raptor-Event say when and what. Check the signature, and reject old timestamps.

Secrets (webhook URLs, tokens) are encrypted and never shown again. The last 50 deliveries, with any errors, are listed on the page.

© 2026 Xena Studios. Pterodactyl is a trademark of its owners; Raptor isn't affiliated with it.