The HopGlass Server collects data from Freifunk networks and processes it to be used in HopGlass, for statistics and other purposes.
Setup is easy:
- Be sure to have nodejs and npm installed
- Install dependencies:
npm install - Start the server:
node hopglass.js [args]
The following arguments can be supplied:
| Argument | Default | Description |
|---|---|---|
| webport | 4000 | webserver port |
| iface | bat0 | the interface to discover nodes on |
| ifaces | bat0 | a comma-seperated list of interfaces to discover nodes on |
| nodeinfointerval | 180 | interval for nodeinfo queries (in seconds, low values can impact mesh performance) |
| statisticsinterval | 60 | interval for statistics and neighbourinfo queries (in seconds, low values can impact mesh performance) |
| collectorport | 45123 | the port the data collector listens on |
| targetip | ff02::2 | IPv6 (usually multicast group) to query |
| targetport | 1001 | the port to query the nodes on |
| Query Location | Description |
|---|---|
| /nodes.json | Meshviewer nodes.json v2 |
| /graph.json | Meshviewer graph.json v1 |
| /raw.json | raw data collected, same as the raw.json save file |
| /hosts | hosts file |
| /metrics | Prometheus metrics (currently gluon-collector-style, might change) |
per node (all with labels hostname and nodeid):
- statistics.clients.total
- statistics.uptime
- statistics.traffic.rx.bytes
- statistics.traffic.mgmt_rx.bytes
- statistics.traffic.tx.bytes
- statistics.traffic.mgmt_tx.bytes
- statistics.traffic.forward.bytes
- statistics.loadavg
- statistics.memory_usage
total values:
- meshnodes_total
- meshnodes_online_total
- total_clients
- total_traffic_rx
- total_traffic_mgmt_rx
- total_traffic_tx
- total_traffic_mgmt_tx
- total_traffic_forward