Since room-assistant is a distributed system it quickly gets difficult to keep track of everything. To help we provide some tools for power users that wish to keep track of all their instance states without having to go to each one manually.
room-assistant provides metrics in the Prometheus format on port
/metrics. This includes some default metrics about the general state of the application, as well as integration-specific metrics that you can explore. You can setup your Prometheus instance to scrape these endpoints like this:
scrape_configs: - job_name: 'room-assistant' metrics_path: '/metrics' scrape_interval: 30s static_configs: - targets: - '<room-assistant-ip1>:6415' - '<room-assistant-ip2>:6415'
If you want to capture the logs of all your instances in a single place you can configure additional log outputs, which will be used alongside the console output you are used to.
room-assistant is capable of piping log output into an Elasticsearch cluster with the logstash format, which makes it searchable from Kibana. On NodeJS installations this logger requires an additional dependency, which you can install with
npm install -g winston-elasticsearch.
|Boolean||Whether room-assistant should send logs to Elasticsearch or not.|
|String||http://localhost:9200||The endpoint under which your Elasticsearch cluster is accessible.|
|Object||Object containing either |
|String||The prefix of the indices that will be created. The indices follow the naming format |
logger: elasticsearch: enabled: true node: http://192.168.0.20:9200
You can send the room-assistant logs to an instance of Grafana Lokiopen in new window, which will allow you to collect and query logs easily from Grafanaopen in new window. On NodeJS installations this logger requires an additional dependency, which you can install with
npm install -g winston-loki.
|Boolean||Whether room-assistant should send logs to Loki or not.|
|String||http://localhost:3100||The endpoint of your Loki instance.|
logger: loki: enabled: true host: http://192.168.0.20:3100