Telegraf Configuration


4. Configure Telegraf for metrics
1) Enable metrics in kcnd

Check if the two options below are enabled in kcnd.conf.

...
METRICS=1
PROMETHEUS=1
...

After kcnd is started, you can check that port 61001 is open.

$ netstat -ntap | grep 61001
tcp        0      0 :::61001        :::*       LISTEN      8989/kcn

2) Set the collection intervals in /etc/telegraf/telegraf.conf

Set the three values below under the [agent] section. If a line is commented out (#), uncomment it.

[agent]
  ...
  interval = "10s"
  round_interval = true
  ...
  flush_interval = "10s"
  ...

3) Create /etc/telegraf/telegraf.d/kaia.conf

Replace <hostname> with the name of your CN in the <org name>-cn format (e.g. abc-cn).

Note

Use the same <hostname> value for the instance of the log shipper in the next step. Otherwise, your node will be split into two identities in the metric and log dashboards.

[global_tags]
nodetype = "cn"
network = "mainnet"
[agent]
# The CN name: <org name>-cn (e.g. `abc-cn`)
hostname = "<hostname>"

[[inputs.procstat]]
  exe = "kcn"

[inputs.prometheus]
urls = [ "http://localhost:61001/metrics" ]

[[outputs.influxdb]]
  urls = ["http://metrics.node.kaia.io:45560"]
  database = "kaia-network-monitoring"

  namedrop = ["kaia_log"]
  skip_database_creation = true

You can check more details on the page below.


If you finish this step, please click the next button > on the right side of this page.