Skip to content

Latest commit

 

History

History
173 lines (121 loc) · 6.14 KB

grafana.md

File metadata and controls

173 lines (121 loc) · 6.14 KB
Raw

Grafana

Grafana is used as a flexible dashboard system that can be used to display the InfluxDB data.

To access the dashboard, go to http://localhost:3001

By default, the credentials are:

  • username : admin
  • password : admin

If the credentials don't work, ask Colin or Candan what they are.

Prerequisites

Before trying to set up the dashboard, we must be sure that data is being written to InfluxDB by Telegraf

Check the telegraf logs:

docker compose logs -f telegraf

You should see:

stack-telegraf-1  | 2023-01-23T14:14:18Z I! Using config file: /etc/telegraf/telegraf.conf
stack-telegraf-1  | 2023-01-23T14:14:18Z I! Starting Telegraf 1.25.0
stack-telegraf-1  | 2023-01-23T14:14:18Z I! Available plugins: 228 inputs, 9 aggregators, 26 processors, 21 parsers, 57 outputs, 2 secret-stores
stack-telegraf-1  | 2023-01-23T14:14:18Z I! Loaded inputs: mqtt_consumer (6x)
stack-telegraf-1  | 2023-01-23T14:14:18Z I! Loaded aggregators: 
stack-telegraf-1  | 2023-01-23T14:14:18Z I! Loaded processors: 
stack-telegraf-1  | 2023-01-23T14:14:18Z I! Loaded secretstores: 
stack-telegraf-1  | 2023-01-23T14:14:18Z I! Loaded outputs: influxdb_v2
stack-telegraf-1  | 2023-01-23T14:14:18Z I! Tags enabled: host=fe44eebea0fe
stack-telegraf-1  | 2023-01-23T14:14:18Z I! [agent] Config: Interval:10s, Quiet:false, Hostname:"fe44eebea0fe", Flush Interval:10s
stack-telegraf-1  | 2023-01-23T14:14:18Z D! [agent] Initializing plugins
stack-telegraf-1  | 2023-01-23T14:14:18Z D! [agent] Connecting outputs
stack-telegraf-1  | 2023-01-23T14:14:18Z D! [agent] Attempting connection to [outputs.influxdb_v2]
stack-telegraf-1  | 2023-01-23T14:14:18Z D! [agent] Successfully connected to outputs.influxdb_v2
stack-telegraf-1  | 2023-01-23T14:14:18Z D! [agent] Starting service inputs
stack-telegraf-1  | 2023-01-23T14:14:18Z I! [inputs.mqtt_consumer] Connected [tcp://mosquitto:1883]
stack-telegraf-1  | 2023-01-23T14:14:18Z I! [inputs.mqtt_consumer] Connected [tcp://mosquitto:1883]
stack-telegraf-1  | 2023-01-23T14:14:18Z I! [inputs.mqtt_consumer] Connected [tcp://mosquitto:1883]
stack-telegraf-1  | 2023-01-23T14:14:18Z I! [inputs.mqtt_consumer] Connected [tcp://mosquitto:1883]
stack-telegraf-1  | 2023-01-23T14:14:18Z I! [inputs.mqtt_consumer] Connected [tcp://mosquitto:1883]
stack-telegraf-1  | 2023-01-23T14:14:18Z I! [inputs.mqtt_consumer] Connected [tcp://mosquitto:1883]
stack-telegraf-1  | 2023-01-23T14:14:28Z D! [outputs.influxdb_v2] Wrote batch of 30 metrics in 6.859893ms
stack-telegraf-1  | 2023-01-23T14:14:28Z D! [outputs.influxdb_v2] Buffer fullness: 0 / 10000 metrics

If you don't see something like this: Wrote batch of 30 metrics in 6.859893ms, something is going wrong. For example, it could be that the InfluxDB token in the file my.env has not been set properly.

This must be solved before proceeding further

InfluxDB data source

Add InfluxDB as a data source:

Note that:

  • We use InfluxDB 2 so we need to use Flux as a query language
  • The URL of the InfluxDB server is http://influxdb:8086. Indeed:
    • Grafana is running in the docker stack. From there, the url of the InfluxDB container is simply influxdb. The port exposed by this container within the stack is 8086. See docker-compose.yml for more information
  • All Auths are uncheched
  • The organization and tokens can be found in your influxdb/config/influx-configs

Save and test your data source and go back to the home menu.

Dashboard creation

Create a new dashboard, and a new panel in this dashboard.

Use the Show bucket sample query and the table view to display the list of buckets in your InfluxDB database:

The _monitoring and _tasks bucket are internal InfluxDB buckets. Our bucket is my-bucket. The name is set in telegraf.conf.

Before going further:

Edit your data source and set my-bucket as a default bucket.

(you can find your data source in the global settings). Then go back to the exploration window.

You can then get the list of measurements within the bucket:

And you can run this query to see the list of fields in the hv measurement

import "influxdata/influxdb/schema"

schema.measurementFieldKeys(
    bucket: "my-bucket",
    measurement: "sensor_1",
)

This should give you two fields, meas1 and meas2

This is expected from the telegraf configuration and from the mqtt logs:

docker exec stack-mosquitto-1 mosquitto_sub -v -t /#
...
/sensor_1/status {"meas1": 0.5439515571875321, "meas2": 0.5439515571875321}
...

Let us now set up our panel to display both meas1 and meas2 in the same plot. For this, use the Simple query template with this query:

from(bucket: "my-bucket")
  |> range(start: v.timeRangeStart, stop:v.timeRangeStop)
  |> filter(fn: (r) => r._measurement == "sensor_1")
  |> filter(fn: (r) => r._field =~ /meas.*/)
  |> aggregateWindow(every: v.windowPeriod, fn: mean, createEmpty: false)
  |> yield(name: "mean")

Note the use of:

  • aggregateWindow: without this, the query will fail if we try and display too many points over a long time range
  • the regex filter: without it, aggregateWindow fails.

We could filter on the field name to get only meas1 or only meas2.

Now save your dashboard.

Exporting and importing dashboards

Dashboards can be exported as json, and imported back later on, even on a different machine.

To do this, click the share button, and export as JSON. Check the box "export for sharing externally", which will allow you to select the datasource on the other system.

To import a dashboard, use the Dashboards menu.

After the import, you won't see any data plotted in the dashboard panels. You need to open the dashboard panels and reselect InfluxDB as a source. Then wait a bit, and you should see data appear.

You may commit your dashboards in ../grafana/dashboards.