Using Beacon
These pages are the work you do with Beacon once it runs, grouped by what Beacon does. Each page is one task: what you need, the steps, how to check the result, and what to read next. New to Beacon? Start with What is Beacon, Try it on your laptop and How Beacon works.
Monitoring and alerts
What Beacon checks, from where, and what it tells your users.
- Monitor a service: probe locations, target sets and modules. Every change is checked, takes effect at once and can be reverted.
- Heartbeats and expiring resources: jobs that must report success often enough, and domains and machines that expire.
- Publish a status page: what the probes agree on, made public, with incidents and maintenance windows.
Cards in the chat for the people on call are set up in Set up the Feishu app.
Agents on call
An alert wakes an agent; the agent reads the evidence, fixes what it can, checks the fix and leaves a note.
- Put an agent on call: the hook that wakes an agent, the evidence in one call, probing on demand and notes.
- Rehearse an outage: the whole loop on Beacon's own drill service, before a real outage.
- Review what agents did: the console, where people see what needs attention, the alerts, silences and maintenance, and every session with its replay.
The agent clients and the other end of a hook are in Integrations.
Machines and the SSH channel
How an agent reaches a server without holding a key, and how to stop it.
- Give agents a machine: turn the channel on, add machines (behind NAT too) and give an agent access.
- Stop an agent, pause everything: every brake, and what each one closes.
The team and permissions
Who may do what, people and their agents alike.
- Invite your team: accounts and roles, agents and their presets, OAuth sign-ins, SSH keys, and moving from tokens.
Every role, preset and scope is listed in Roles, presets and scopes.