bassista / influxdbroute

Using Apache Camel to Store and Visualize Data From The Things Network using Grafana and InfluxDB

Using Apache Camel to Store and Visualize Data From The Things Network using Grafana and InfluxDB

This example was inspired by Laurens Slats example

https://www.thethingsnetwork.org/labs/story/store-and-visualize-data-using-influxdb-and-grafana

This example demonstrates how you can use Apache Camel with Spring Boot to consume messages sent to The Things Network by motes (devices) and send them to an InfluxDB database.

The example can be run locally but this requires an existing InfluxDB instance. There is also a description of how to run the example in Openshift including how to deploy InfluxDB and Grafana so that a full end to end solution can be demonstrated.

Building

The example can be built with

mvn clean install

Running locally

The example can be run locally (replacing the exported values with those applicable to your TTN application and your Influx DB).

export REGION=eu #this is TTN region
export APP_ID=<<your app id>>
export ACCESS_KEY=<<your access key>>
export DEVICE_ID=<<your device id>>
export INFLUXDB_DB=<<the influx db name>>
export INFLUXDB_URL='http://<<influx db host>>:<<influx db port'

mvn spring-boot:run

Running the example in OpenShift

This code base was generated using the FIS 2.0 maven archetypes https://access.redhat.com/documentation/en-us/red_hat_jboss_fuse/6.3/html/fuse_integration_services_2.0_for_openshift/ . Obviously to deploy to Openshift an Openshift environment is needed. The easiest way to do this in a local environment is to use minishift https://access.redhat.com/documentation/en-us/red_hat_jboss_fuse/6.3/html/fuse_integration_services_2.0_for_openshift/get-started-dev#get-started-prereq .

Firstly you will need to deploy an InfluxDB server and a Grafana instance into your Openshift instance.

git clone https://github.com/feedhenry/sync-metrics-openshift
export INFLUXDB_STORAGE=1Gi
cd sync-metrics-openshift/
oc new-app -p STORAGE_SIZE="$INFLUXDB_STORAGE" -f ./influxdb-template.yaml
export GRAFANA_STORAGE=2Gi 
oc new-app -p STORAGE_SIZE="$GRAFANA_STORAGE" -f ./grafana-template.yaml

Then use oc get pod to find out the name of the influxdb pod. Then obtain a remote shell to that pod oc rsh <<pod name>> bash and then

influx
  Connected to http://localhost:8086 version 1.5.1
  InfluxDB shell version: 1.5.1
> CREATE DATABASE ttndb
> USE ttndb
  Using database ttndb
> create retention policy "default" on ttndb duration 0d replication 1 default

The example can be built and run on OpenShift using a single goal (if you experience issues with missing image streams then follow instructions in the section 'Importing Image Streams'. Update the file src/main/fabric8/deployment.yml with values applicable to your TTN application, device:

mvn fabric8:deploy

When the example runs in OpenShift, you can use the OpenShift client tool to inspect the status

To list all the running pods:

oc get pods

Then find the name of the pod that runs this quickstart, and output the logs from the running pods with:

oc logs <name of pod>

You can also use the openshift web console to manage the running pods, and view logs and much more.

Setting up a Grafana Dashboard

Then follow the Grafana documentation to setup a dashboard http://docs.grafana.org/features/datasources/influxdb/#using-influxdb-in-grafana . When setting up the InfluxDB datasource use the following for the properties:

• The name of the influx db ttndb
• The username can be left blank
• The password can be left blank
• Leave the 'Access' mode as 'proxy'
• For the URL use the Openshift service name http://influxdb-svc:8086

The measurement name is 'temperature' and the value name is 'temperature' i.e. the Grafana query should look something like

SELECT mean("temperature") FROM "temperature" WHERE $timeFilter GROUP BY time($interval) fill(none)

Importing Image Streams

Import the Fuse image streams:

oc create -f https://raw.githubusercontent.com/jboss-fuse/application-templates/GA/fis-image-streams.json