Get started with logging on Kubernetes using Vector and send your logs to logfire. Start logging in just 2 minutes!

Installation Methods

There are two primary methods to install Vector. In this guide, we will cover the Kustomize method:

Installing Vector on Kubernetes using kubectl

kubectl is the Kubernetes command-line tool. While Helm is commonly used for package management in Kubernetes, kubectl can also be used to deploy applications and manage configurations. The following instructions will guide you through installing Vector in both the Agent and Aggregator roles using kubectl.

Agent

The Vector Agent is a key component that allows you to collect data from your sources and then forward it to various destinations using sinks.

1. Define Vector’s Namespace

Running Vector in its dedicated namespace is recommended for isolation and management. In this guide, we will use vector as the namespace, but you can choose any appropriate name:

kubectl create namespace --dry-run=client -o yaml vector > namespace.yaml

2. Prepare your kustomization file

This kustomization configuration deploys Vector as an Agent. You can review the full default configuration here. For deeper insights into configurations, visit the configuration docs page.

cat <<-'KUSTOMIZATION' > kustomization.yaml
---
apiVersion: kustomize.config.k8s.io/v1beta1
kind: Kustomization

# Override the namespace of all of the resources we manage.
namespace:
  vector

# Override the Vector image to pin the version used.
images:
  - name: timberio/vector
    newName: timberio/vector
    newTag: 0.31.0-distroless-libc

    # The namespace previously created to keep the resources in.
resources:
  - namespace.yaml
  - rbac.yaml
  - https://github.com/vectordotdev/vector/distribution/kubernetes/vector-agent

# configMapGenerator:
#   - behavior: replace
#     files:
#       - vector-agent.toml
#     name: vector

patches:
  # Prevent data dir read-only volume mount issue: https://github.com/vectordotdev/helm-charts/issues/226
  # Set volumeMount readOnly: false
  # Also, add vector SA token into an ENV variable, Vector will use it to connect to metrics-server
  - patch: |-
      - op: replace
        path: /spec/template/spec/containers/0/volumeMounts/3/readOnly
        value: false
      - op: add
        path: /spec/template/spec/containers/0/env/-
        value:
          name: SERVICE_ACCOUNT_TOKEN
          valueFrom:
            secretKeyRef:
              name: vector-service-account
              key: token
    target:
      group: apps
      version: v1
      kind: DaemonSet
      name: vector
KUSTOMIZATION

Save the following config as rbac.yaml to allow Vector to access the Metrics Server:

cat <<-'RBAC' > rbac.yaml
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRole
metadata:
  name: vector-metrics
  labels:
    app.kubernetes.io/name: vector
    app.kubernetes.io/instance: vector
    app.kubernetes.io/component: Agent
rules:
  - apiGroups:
      - "metrics.k8s.io"
    resources:
      - nodes
      - pods
    verbs:
      - list
---
apiVersion: rbac.authorization.k8s.io/v1
kind: ClusterRoleBinding
metadata:
  name: vector-metrics
  labels:
    app.kubernetes.io/name: vector
    app.kubernetes.io/instance: vector
    app.kubernetes.io/component: Agent
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: ClusterRole
  name: vector-metrics
subjects:
  - kind: ServiceAccount
    name: vector
    namespace: vector
---
apiVersion: v1
kind: Secret
metadata:
  name: vector-service-account
  annotations:
    kubernetes.io/service-account.name: "vector"
type: kubernetes.io/service-account-token
RBAC

3. Verify your kustomization file

Ensure that the kustomization configuration is correctly set up:

kubectl kustomize

4. Install Vector

Apply the configurations to deploy Vector:

kubectl apply -k .

5. Tail Vector logs

To check the logs of the Vector instance:

kubectl logs -n vector daemonset/vector

6. Update kustomization.yaml

Firstly, open your existing kustomization.yaml file for editing. You can use any text editor or command-line editor for this.

Append the following configMapGenerator block to your kustomization.yaml:

configMapGenerator:
  - name: vector
    behavior: replace
    files:
      - vector-agent.toml

This will generate a new Kubernetes ConfigMap named vector from the vector-agent.toml file and replace any existing ConfigMap with the same name.

7. Download the New Vector Configuration

Run the following command to download your specific Vector configuration:

wget -O vector-agent.toml \
   https://api.logfire.ai/configuration/download/vector-toml/kubernetes/$SOURCE_TOKEN

Note: Replace $SOURCE_TOKEN with your actual source token which can be found in your sources.

Ensure the vector-agent.toml file is in the same directory as your kustomization.yaml or provide the appropriate path in the kustomization.yaml.

8. Reload & Restart

To apply the changes and refresh Vector, execute:

kubectl apply -k .
kubectl rollout restart -n vector daemonset/vector

This command instructs Kubernetes to process the kustomization.yaml file, generate the required resources, and update the cluster accordingly. Since you’ve defined the behavior as replace in the configMapGenerator, it will replace the existing ConfigMap with the newly defined one.

After this step, Vector should pick up the new configuration. Ensure to monitor logs and the status of the Vector pods to confirm the successful application of the new configuration.

Check your logs in Logfire -> Live tail.

Warning: Kubernetes version 1.15 or higher is required.

Info: Kubelet certificate needs to be signed by the cluster Certificate Authority. If using Minikube, consider disabling certificate validation by modifying the metrics-server DaemonSet and vector-agent.toml.

Need help? Reach out at support@logfire.ai . We’re here to assist!

Next Steps

Once you have completed the above steps, you can continue to set up Vector to gather logs and send them to logfire. Make sure to monitor the logs and ensure that the Vector setup is working seamlessly with logfire.

For more advanced setups or for using the Helm method, please refer to our detailed documentation or community forums.