KNX Prometheus Exporter

The KNX Prometheus Exporter is a small bridge to export values measured by KNX sensors to Prometheus. It takes the values either from cyclic sent GroupValueWrite telegrams and can request values itself using GroupValueRead telegrams.

The data stored in Prometheus can be visualized with Grafana. This could look like this screenshot:

• KNX Prometheus Exporter
  • Installing
    • macOS
    • Docker
    • Other
  • Usage
    • Converting the ETS 5 Group Export to a configuration
    • Preparing the configuration
      • The Connection section
        • The RouterConfig
        • The TunnelConfig
      • The MetricsPrefix
      • The ReadStartupInterval
      • The AddressConfigs section
    • Running the exporter
    • Running the exporter using docker
  • Exported metrics
  • Health Check Endpoints
  • Contributing
    • Renamed branch to main
  • License
  • Maintainer

Installing

macOS

Using brew tap:

brew tap chr-fritz/tap
brew install chr-fritz/tap/knx-exporter

Docker

docker pull quay.io/chrfritz/knx-exporter

Other

Download the latest release and place it in your $PATH.

Usage

Converting the ETS 5 Group Export to a configuration

The KNX Prometheus Exporter will only export the values from configured group addresses. A good starting point is to export the group addresses from ETS 5 into the XML format and convert them.

Please refer to the KNX Documentation "Group Address & Export" for more information about how to export the group addresses to XML.

With the exported group addresses you can call the knx-exporter and convert them:

knx-exporter convertGA [SOURCE] [TARGET]

You must replace [SOURCE] with the path to your group address export file. [TARGET] is the path where the converted configuration should be stored.

Preparing the configuration

Converting the group addresses is a good starting point for preparing the actual configuration. The simplest fully configuration will look like this:

Connection:
    Type: "Tunnel"
    Endpoint: "192.168.1.15:3671"
    PhysicalAddress: 2.0.1
MetricsPrefix: knx_
ReadStartupInterval: 200ms
AddressConfigs:
    0/0/1:
        Name: dummy_metric
        DPT: 1.*
        Export: true
        MetricType: "counter"
        ReadStartup: true
        ReadActive: true
        MaxAge: 10m
        Comment: dummy comment
        Labels:
            room: office

In the next sections we will describe the meanings for every statement.

The Connection section

The Connection section contains all settings about how to connect to your KNX system and how the KNX Prometheus Exporter will identify itself within it. It has three properties:

• Type This defines the connection type to your KNX system. It can be either Router or Tunnel.
• Endpoint This defines the ip address or hostname including the port to where the KNX Prometheus Exporter should open the connection. In case of you are using Router in Type the default might be 224.0.23.12:3671.
• PhysicalAddress This defines the physical address of how the KNX Prometheus Exporter will identify itself within your KNX address.
• RouterConfig This defines additional
• TunnelConfig contains some specific configurations if Type is Tunnel

The RouterConfig

• RetainCount specifies how many sent messages to retain. This is important for when a router indicates that it has lost some messages. If you do not expect to saturate the router, keep this low.
• Interface specifies the name of the network interface used to send and receive KNXnet/IP packets. If the interface is nil, the system-assigned multicast interface is used.
• MulticastLoopbackEnabled specifies if Multicast Loopback should be enabled.
• PostSendPauseDuration specifies the pause duration after sending. 0 means disabled. According to the specification, we may choose to always pause for 20 ms after transmitting, but we should always pause for at least 5 ms on a multicast address.

The TunnelConfig

• ResendInterval is the interval with which requests will be resent if no response is received.
• HeartbeatInterval specifies the time interval which triggers a heartbeat check.
• ResponseTimeout specifies how long to wait for a response.
• SendLocalAddress specifies if local address should be sent on connection request.
• UseTCP configures whether to connect to the gateway using TCP.

The MetricsPrefix

The MetricsPrefix defines a single string that will be added to all your exported metrics. The format must be compliant with the prometheus metrics names.

The ReadStartupInterval

It is possible to send GroupValueRead telegrams to specific group addresses at startup. Sending out all GroupValueRead telegrams at once on startup can overwhelm the KNX bus. The ReadStartupInterval defines the interval between the GroupValueRead telegrams sent out at startup. If not specified, ReadStartupInterval is set to 200ms by default.

The AddressConfigs section

The AddressConfigs section defines all the information about the group addresses which should be exported to Prometheus. It contains the following structure for every exported group address.

  0/0/1:
      Name: dummy_metric
      DPT: 1.*
      Export: true
      MetricType: "counter"
      ReadStartup: true
      ReadActive: true
      MaxAge: 10m
      Comment: dummy comment
      ReadType: WriteOther
      ReadAddress: 0/0/2
      ReadBody: [ 0x1 ]
      Labels:
          room: office

The first line 0/0/1 defines the group address that should be exported. It can be written in all three valid forms:

• Two layers: 0/1
• Three layers: 0/0/1
• Free structure 123

You can find more information about the formatting of group addresses here: cemi@v0.0 NewGroupAddrString

Next it defines the actual information for a single group address:

• Name is the short name of the exported metric. The format must be compliant with the prometheus metrics names.
• DPT The knx data point type. This defines how the KNX Prometheus Exporter will interpret the payload of received telegrams. All available data point types can be found here: knx dpt
• Export Can be either true or false. Allows to disable exporting the group address as value.
• MetricType defines the type of the exported metric. Can be either counter or gauge. See Prometheus documentation counter vs. gauge for more information about it.
• ReadStartup can either be true or false. If set to true the KNX Prometheus Exporter will send a GroupValueRead telegram to the group address to actively ask for a new value once after startup. In contrast to ReadActive this sends out a GroupValueRead telegram at startup once.
• ReadActive can either be true or false. If set to true the KNX Prometheus Exporter will send a GroupValueRead telegram to the group address to active ask for a new value if the last received value is older than MaxAge.
• ReadType defines the type how to trigger the read request. Possible Values are GroupRead and WriteOther. Default is GroupRead.
• ReadAddress defines the group address to which address a GroupWrite request should be sent to initiate sending the data if ReadType is set to WriteOther.
• ReadBody is a byte array with the content to sent to ReadAddress if ReadType is set to WriteOther.
• MaxAge defines the maximum age of a value until the KNX Prometheus Exporter will send a GroupValueRead telegram to active request a new value for the group address. This setting will be ignored if ReadActive is set to false.
• Comment a short comment for the group address. Will be also exported as comment within the Prometheus metrics.
• Labels are additional information for a specific time series. A common usage of labels could be a label room which identifies the room for a metric current_temperature.

Running the exporter

To run the metrics export just run the following command:

knx-exporter run -f [CONFIG-FIlE]

You must replace [CONFIG-FILE] with the path to your configuration file that you prepared in the previous step. After starting the exporter you can open http://localhost:8080/metrics to view the exported metrics.

Running the exporter using docker

It is also possible to run the KNX Exporter using docker. For this just run the following command:

docker run -p 8080:8080 -v [CONFIG-FILE]:/etc/knx-exporter/ga-config.yaml quay.io/chrfritz/knx-exporter

As before you must replace [CONFIG-FILE] with the path to your configuration file that you prepared in the previous step then you can open http://localhost:8080/metrics to view the exported metrics.

Exported metrics

Beside exported metrics from KNX group addresses it exports some additional metrics. This metrics can be grouped into three groups:

1. KNX Message Metrics: Three counter metrics which counts the number of
   • received messages knx_messages{direction="received",processed="false"}
   • processed received messages knx_messages{direction="received",processed="true"} and
   • sent messages knx_messages{direction="sent",processed="true"}.
2. HTTP Metrics: Counts the processed number of successfully and failed http requests. All metrics starts with promhttp_.
3. GoLang Metrics: These are metrics that indicate some health information about memory, cpu usage, threads and other GoLang internal information.

Health Check Endpoints

The KNX Prometheus Exporter provides endpoints for liveness and readiness checks like they were recommended in environments like Kubernetes. They are reachable under /live and /ready. With /live?full=1 and /ready?full=1 they also print the status of every single check.

Contributing

1. Fork it
2. Download your fork to your PC (git clone https://github.com/your_username/knx-exporter && cd knx-exporter)
3. Create your feature branch (git checkout -b my-new-feature)
4. Make changes and add them (git add .)
5. Commit your changes (git commit -m 'Add some feature')
6. Push to the branch (git push origin my-new-feature)
7. Create new pull request

Renamed branch to main

The name of default branch recently changed from master to main. If you have a local working copy of this repository you can rename it using the following commands:

git branch -m master main
git fetch origin
git branch -u origin/main main
git remote set-head origin -a

License

The KNX Exporter is released under the Apache 2.0 license. See LICENSE

Maintainer

Christian Fritz (@chrfritz)