Operate

A healthy Junkmesh node is a boring one: lid closed, plugged into power and Ethernet, forgotten in a cupboard. These pages are for the moments in between — checking on it, using the storage it gives you, and watching the whole cluster from one dashboard.

• Running a node

  The owner's manual: health checks, updates, taking a node offline, backing up its identity, laptop-specific care.

• Using the storage

  S3 credentials, rclone / AWS CLI / restic recipes, and client-side encryption so node hosts only ever hold ciphertext.

• Metrics & monitoring

  Every node serves Prometheus metrics, a JSON status API and mesh-native discovery on port 3904. Point Prometheus, an OpenTelemetry Collector or New Relic at one node and it finds the whole cluster.

The one command worth memorising:

$ curl -s localhost:3904/api/v1/status    # everything about this node, as JSON