victoriametrics-vmstorage

Download this Image

The image is available on cgr.dev:

docker pull cgr.dev/chainguard/victoriametrics:latest

VictoriaMetrics can be deployed as a single node (all image tags without a "-cluster" suffix) or as a distributed cluster. See the github repository for reference: https://github.com/VictoriaMetrics/VictoriaMetrics/tree/master as well as their public facing documentation: https://docs.victoriametrics.com/

In clustered mode, the VictoriaMetrics application itself is broken into 3 different services:

• vmstorage - stores the raw data and returns the queried data on the given time range for the given label filters
• vminsert - accepts the ingested data and spreads it among vmstorage nodes according to consistent hashing over metric name and all its labels
• vmselect - performs incoming queries by fetching the needed data from all the configured vmstorage nodes
• vmalert - executes a list of given alerting / monitoring rules against Prometheus HTTP API

This image group also contains several standalone services (that DO vary between clustered and single mode):

• vmagent - A lightweight metrics collection agent.
• vmauth - A service that provides authentication and authorization for VictoriaMetrics services. It acts as HTTP proxy, which can authorize, route and load balance requests across VictoriaMetrics components or any other HTTP backends.

Usage

You can install the VictoriaMetrics components using Helm. Below are the Helm install commands for each component:

Install vmagent

helm install vmagent victoria-metrics-agent --repo https://victoriametrics.github.io/helm-charts/ \
  --namespace victoriametrics \
  --set image.repository=cgr.dev/chainguard-private/victoriametrics-vmagent \
  --set image.tag=latest

Install vmauth

helm install vmauth victoria-metrics-auth --repo https://victoriametrics.github.io/helm-charts/ \
  --namespace victoriametrics \
  --set image.repository=cgr.dev/chainguard-private/victoriametrics-vmauth \
  --set image.tag=latest

Note: You will need to set the config and for sample testing purpose, you can take inspiration from this example:

cat <<YAML | kubectl apply -f -
apiVersion: v1
kind: ConfigMap
metadata:
  name: vmauth-config
  namespace: victoriametrics
data:
  auth.yml: |
    users:
      - username: "user1"
        password: "pass1"
        url_prefix: "http://localhost:8428/"
      - username: "user2"
        password: "pass2"
        url_prefix: "http://localhost:8429/"
    unauthorized_user:
      url_prefix: "http://localhost:8428/"
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: vmauth
  namespace: victoriametrics
spec:
  selector:
    matchLabels:
      app: vmauth
  template:
    metadata:
      labels:
        app: vmauth
    spec:
      containers:
      - name: vmauth
        image: cgr.dev/chainguard-private/victoriametrics-vmauth:latest
        args:
          - "-auth.config=/config/auth.yml"
        ports:
        - containerPort: 8427
        volumeMounts:
        - name: vmauth-config
          mountPath: /config
      volumes:
      - name: vmauth-config
        configMap:
          name: vmauth-config
YAML

Install vminsert, vmselect, vmstorage

helm install vminsert victoria-metrics-cluster --repo https://victoriametrics.github.io/helm-charts/ \
  --namespace victoriametrics-cluster \
  --set vminsert.image.repository=cgr.dev/chainguard-private/victoriametrics-vminsert \
  --set vminsert.image.tag=latest \
  --set vminsert.image.variant="" \
  --set vmselect.image.repository=cgr.dev/chainguard-private/victoriametrics-vmselect \
  --set vmselect.image.tag=latest \
  --set vmselect.image.variant="" \
  --set vmstorage.image.repository=cgr.dev/chainguard-private/victoriametrics-vmstorage \
  --set vmstorage.image.tag=latest \
  --set vmstorage.image.variant=""

Note: We use vmstorage.image.variant="" because our images are alredy coming with cluster suffix, so we no longer need to specify it. For further reference, refer to official docs

Install vmalert

You can use this example and override the image to use chainguard images like shown below. Note the server.config is your alerting rules, so you will have to customize those as per your own needs, as well as other things that are pertinent to your own setup.

cat <<EOF > /tmp/helminstall
server:
  datasource:
    url: "your PROMETHEUS HTTP API server here"
  image:
    repository: cgr.dev/chainguard-private/victoriametrics-vmalert
    tag: latest
  notifier:
    alertmanager:
      url: http://localhost:9093
  config:
    alerts:
      groups:
        - name: common
          rules:
            - alert: instanceIsDown
              for: 1m
              expr: up == 0
              labels:
                severity: critical
              annotations:
                summary: "{{ $labels.job }} instance: {{$labels.instance }} is not up"
                description: "Job {{ $labels.job }}  instance: {{$labels.instance }} is not up for the last 1 minute"
EOF

helm install vmalert victoria-metrics-alert --namespace victoriametrics \
--create-namespace --repo https://victoriametrics.github.io/helm-charts/ \
-f /tmp/helminstall

helm install vminsert victoria-metrics-cluster --repo https://victoriametrics.github.io/helm-charts/ \
  --namespace victoriametrics-cluster \
  --set server.image.repository=cgr.dev/chainguard-private/victoriametrics-vmalert \
  --set server.image.tag=latest \
  --set server.image.variant=""
  --set server.config="{...}"