Hello,
I configured grafana agent on a linux EC2 instance and I used metrics to create a dashboard.
I would know just a couple of things about grafana agent:
-
in my grafana-agent.yaml I set “scrape interval” to 1m.
However, in Grafana, if I use 1m as rate interval (e.g this query avg (sum by (cpu) (rate(node_cpu_seconds_total{mode!=“idle”}[1m]))) * 100) I see “No data”.
If I write 30s as scrape interval, I have data as query result.
So, my question is: is this behavior normal? Do I have to set 30s to use 1m range in query operator rate? -
Grafana Agent scrapes a lot of metrics: I need just 10 (more or less) metrics (CPU, RAM, Disk, Load) but I noticed that Grafana Agent scrapes a lot of metrics: can I filter them?
Should I specify in configuration file metrics that I want?
Right now, here is my grafana-agent.yaml (don’t pay attention to logs section)
server:
log_level: warn
metrics:
global:
scrape_interval: 30s
#wal_directory: ‘/var/lib/grafana-agent’
remote_write:
- url: https://mimir.subcom.it:443/api/v1/push
headers:
X-Scope-OrgID: subcomgagent
configs:
logs:
configs:
- name: default
clients:- url: https://loki.mimir.subcom.it:443/loki/api/v1/push
tenant_id: diego31
positions:
filename: /tmp/positions.yaml
scrape_configs: - job_name: secure
pipeline_stages:- regex:
expression: ‘(?P\w{3} ( |1|2|3)\d{1} \d{2}:\d{2}:\d{2}) .*’ - timestamp:
source: timestamp
format: “Jan 2 15:04:05” #May 7 03:35:01 ansible systemd: Started Session 598194 of user root
location: “Europe/Rome”
static_configs: - targets: [localhost]
labels:
job: secure
path: /var/log/secure
host: ansible
- regex:
- url: https://loki.mimir.subcom.it:443/loki/api/v1/push
integrations:
agent:
enabled: true
node_exporter:
enabled: true
include_exporter_metrics: true
disable_collectors:
- “mdadm”
Thank you in advance