0d93aa9081 | ||
---|---|---|
.. | ||
README.md | ||
prometheus.go |
README.md
prometheus
Categraf simply wrapped prometheus agent mode fuction, which means you can scrape metrics like prometheus do and get full support with prometheus plugins, such as service discovery and relabel.
For more details, see the official docs:
Configuration
An example to scrape kube-apiserver and core-dns metrics . more examples click here
How to create token
- crate token
kubectl apply -f auth.yaml
, replace CATEGRAF_NAMESPACE with your own in auth.yaml
### auth.yaml
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
annotations: {}
labels:
app: n9e
component: categraf
name: categraf-role
rules:
- apiGroups: [""]
resources:
- nodes
- nodes/metrics
- services
- endpoints
- pods
verbs: ["get", "list", "watch"]
- apiGroups:
- extensions
- networking.k8s.io
resources:
- ingresses
verbs: ["get", "list", "watch"]
- nonResourceURLs: ["/metrics", "/metrics/cadvisor"]
verbs: ["get"]
---
apiVersion: v1
kind: ServiceAccount
metadata:
annotations: {}
labels:
app: n9e
component: categraf
name: categraf-serviceaccount
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
annotations: {}
labels:
app: n9e
component: categraf
name: categraf-rolebinding
roleRef:
apiGroup: rbac.authorization.k8s.io
kind: ClusterRole
name: categraf-role
subjects:
- kind: ServiceAccount
name: categraf-serviceaccount
namespace: ${CATEGRAF_NAMESPACE}
---
- get token
Recommended Strongly: Scraping in cluster, token will be auto mount into pod with path
/var/run/secrets/kubernetes.io/serviceaccount/token
, you do not need to care about it. Replace all Vars with your own in filek8s/in_cluster_scrape.yaml
.
Scraping out of cluster, you can get token with this way and save it to file, then fill bearer_token_file
in file k8s/scrape_with_token.yaml
secrets=$(kubectl get serviceaccount categraf-serviceaccount -o jsonpath={.secrets[].name})
kubectl get secrets ${secrets} -o jsonpath={.data.token} | base64 -d
k8s/scrape_with_cafile.yaml
and k8s/scrape_with_kubecofnig.yaml
is recommended only if you are proficient in
X509 client certs.