Influxdb + Telegraf + Grafana

1

Hello,

I’m currently using Prometheus with Grafana to capture server metrics and display on a nice Grafana Dashboard. However some dashboards you can import are built round Influxdb and Telegraf.

So I’ve installed Influx DB and then installed the Telegraf agent on a couple of VMs and the data is not being pulled into Grafana, all I see is the server name.

How can I check if the InfluxDB is receiving the data from the 2 servers?

Telegraf log file:

2019-09-18T09:41:00Z D! [outputs.influxdb] Wrote batch of 17 metrics in 27.28973ms
2019-09-18T09:41:00Z D! [outputs.influxdb] Buffer fullness: 6 / 10000 metrics
2019-09-18T09:41:04Z D! [agent] Stopping service inputs
2019-09-18T09:41:04Z D! [agent] Input channel closed
2019-09-18T09:41:04Z I! [agent] Hang on, flushing any cached metrics before shutdown
2019-09-18T09:41:04Z D! [outputs.influxdb] Wrote batch of 6 metrics in 1.72249ms
2019-09-18T09:41:04Z D! [outputs.influxdb] Buffer fullness: 0 / 10000 metrics
2019-09-18T09:41:04Z D! [agent] Closing outputs
2019-09-18T09:41:04Z D! [agent] Stopped Successfully

Thanks

2

Looks like at least some data is being written successfully.

You can explore the data in Grafana just by creating a new panel and editing the query, at a bare minimum you need to click on measurement and field(value) and select one of the choices.

Another option is to use the influx command to take a look from a terminal:

$ influx
Connected to http://localhost:8086 version 1.7.8
InfluxDB shell version: 1.7.8
> use telegraf
Using database telegraf
> show measurements
name: measurements
name
----
cpu
> select * from cpu where time > now() - 1m;
3

Measurements are coming through to Influx, I just get get Grafana to show them:

1569078730000000000 cpu-total svr-confluence 0           0                99.24736578011152 0.05017561465144254 0         0          0.050175614651469275 0           0.3010536879083878  0.35122930256092666
1569078730000000000 cpu0      svr-confluence 0           0                99.39698492464723 0                   0         0          0.10050251256267649  0           0.20100502512521015 0.30150753768781524
1569078730000000000 cpu1      svr-confluence 0           0                99.19759277875312 0.10030090270840554 0         0          0                    0           0.4012036108343972  0.3009027081249427
>
>
>
>
> show measurements
name: measurements
name
----
cpu
disk
diskio
kernel
mem
processes
swap
system
>

Lists in Grafana

image
image.png991×757 35.2 KB

I just can’t seem to get the metrics displayed, using any query.

4

Pick cpu and then on the SELECT line click on value in the field(value) box, there should be another dropdown with the available field names. After you pick one I believe at that point you should see some data.

5

Hello guys,
I have similar insallation but when I use “show measurements” function in the terminal I have no response.
Here is the result, does it mean no data is registered in the database ?

SHOW MEASUREMENTS
SHOW MEASUREMENTS

Here is the log of telegraf:
2019-11-29T22:11:00Z I! [inputs.mqtt_consumer] Connected [tcp://127.0.0.1:1883]
2019-11-29T22:33:16Z I! [inputs.mqtt_consumer] Connected [tcp://127.0.0.1:1883]
2019-11-29T22:34:57Z I! Loaded inputs: mqtt_consumer
2019-11-29T22:34:57Z I! Loaded aggregators:
2019-11-29T22:34:57Z I! Loaded processors:
2019-11-29T22:34:57Z I! Loaded outputs: influxdb
2019-11-29T22:34:57Z I! Tags enabled: host=VM-BrokerABBFR
2019-11-29T22:34:57Z I! [agent] Config: Interval:10s, Quiet:false, Hostname:“VM-BrokerABBFR”, Flush Interval:10s
2019-11-29T22:34:57Z I! [inputs.mqtt_consumer] Connected [tcp://127.0.0.1:1883]
2019-12-01T00:02:26Z I! [agent] Hang on, flushing any cached metrics before shutdown
2019-12-01T00:03:24Z I! Loaded inputs: mqtt_consumer
2019-12-01T00:03:24Z I! Loaded aggregators:
2019-12-01T00:03:24Z I! Loaded processors:
2019-12-01T00:03:24Z I! Loaded outputs: influxdb
2019-12-01T00:03:24Z I! Tags enabled: host=VM-BrokerABBFR
2019-12-01T00:03:24Z I! [agent] Config: Interval:10s, Quiet:false, Hostname:“VM-BrokerABBFR”, Flush Interval:10s
2019-12-01T00:03:24Z I! [inputs.mqtt_consumer] Connected [tcp://127.0.0.1:1883]
2019-12-01T21:09:45Z I! [agent] Hang on, flushing any cached metrics before shutdown
2019-12-01T21:09:51Z I! Loaded inputs: mqtt_consumer
2019-12-01T21:09:51Z I! Loaded aggregators:
2019-12-01T21:09:51Z I! Loaded processors:
2019-12-01T21:09:51Z I! Loaded outputs: influxdb
2019-12-01T21:09:51Z I! Tags enabled: host=VM-BrokerABBFR
2019-12-01T21:09:51Z I! [agent] Config: Interval:10s, Quiet:false, Hostname:“VM-BrokerABBFR”, Flush Interval:10s
2019-12-01T21:09:51Z I! [inputs.mqtt_consumer] Connected [tcp://127.0.0.1:1883]
2019-12-01T21:13:46Z I! [agent] Hang on, flushing any cached metrics before shutdown
2019-12-01T21:13:49Z I! Loaded inputs: mqtt_consumer
2019-12-01T21:13:49Z I! Loaded aggregators:
2019-12-01T21:13:49Z I! Loaded processors:
2019-12-01T21:13:49Z I! Loaded outputs: influxdb
2019-12-01T21:13:49Z I! Tags enabled: host=VM-BrokerABBFR
2019-12-01T21:13:49Z I! [agent] Config: Interval:10s, Quiet:false, Hostname:“VM-BrokerABBFR”, Flush Interval:10s
2019-12-01T21:13:49Z I! [inputs.mqtt_consumer] Connected [tcp://127.0.0.1:1883]

6

Doesn’t look like anything is being collected, if you run Telegraf with --debug it will log how many metrics are sent. I also recommend enabling the internal input plugin, which will send metrics about Telegraf activity to InfluxDB.