Running Axiom on Kubernetes

To get Axiom running on Kubernetes use our Helm chart.

The following documentation assumes some knowledge of working with Kubernetes and Helm charts.

If you would like further help or installation in a different environment, please contact support.

Add the Axiom Helm Repository

$ helm repo add axiom https://axiomhq.github.io/axiom-helm-charts
$ helm repo update

Create `values.yaml`

The latest version of values.yaml can be found in the github repository.

# Values file for axiom.
# This is a YAML-formatted file.

# License token (as provided by Axiom support)
licenseToken: "" # if not supplied, you'll end up with the free plan

# [required] Postgres endpoint, should be in the following format:
# "postgresql://host:port/dbname?user=[username]&password=[pwd]&sslmode=require"
# the postgres user provided must be able to create new tables in the DB
postgresUrl: ""

# Redis URL, should be in the following format:
# "redis://:<password (optional)>@<address>/<db (optional)>"
# Examples:
#   redis://redis.example.org:6379
#   redis://:qwerty@localhost:6379/1
redisUrl: ""

# If you want, you can set a different Redis instance as cache registry.
# If you leave this unset, it will use the above redisUrl as cache registry.
cacheRegistryUrl: ""

# Lambda query runner URL, you can set to the query fn ALB URL
# Its usage overrides axiomdb-query-fn deployment
lambdaUrl: ""

# [required] URL specifying how to send emails:
# "sendgrid://apikey?from=noreply@domain.org&fromName=name"
# "smtp://username:password@address:port?from=origin-email&fromName=Axiom&insecure=true"
emailUrl: ""

# URL of this instance, email notifications will use this to link back
# "https://my-axiom.company.org"
externalUrl: ""

# Definition of storage
storage:
  primary:
    # [required] Set to a URI based on the cloud you're using, we support azure / aws / do / gcp
    # Azure: "blob://container"
    # AWS: "s3://bucketName"
    # DigitalOcean: "spaces://bucketName"
    # GCP: "gcs://bucketName"
    # The URI can optionally be suffixed by a path prefix (for example "s3://bucketName/path/prefix")
    #uri: "s3://axiom"

    # AWS-specific properties
    #awsRegion: 'us-east-1'
    #awsAccessKeyID: ''
    #awsSecretAccessKey: ''

    # Azure-specific properties
    #azureStorageAccount: ''
    #azureStorageAccessKey: ''

    # DigitalOcean-specific properties
    #spacesRegion: 'nyc3'
    #spacesKey: ''
    #spacesSecret: ''

    # GCP-specific properties
    #googleApplicationCredentials: '{
    #  "type": "service_account",
    #  "project_id": "...",
    #}'

  fallback:
    # Fallback storage is used when primary is not reachable
    # use the same format as primary storage including credentials
    uri: ""

ingress:
  enabled: false
  annotations:
    {}
    # kubernetes.io/ingress.class: nginx
    # kubernetes.io/tls-acme: "true"
  hosts:
    - host: my-axiom.local
      paths: ["/*"]
  tls: []
  #  - secretName: my-axiom-tls
  #    hosts:
  #      - my-axiom.local

# Settings for axiom-core
core:
  replicas: 3
  resources:
    requests:
      cpu: 250m
      memory: 256Mi
    # limits:
    #   cpu: 500m
    #   memory: 1Gi
  podAnnotations: {}
  nodeSelector: {}
  tolerations: []
  affinity: {}
  extraEnvs: [] # [{ name: AXIOM_XYZ, value: "xyz" }, ... ]

# Settings for axiom-db
db:
  replicas: 3
  resources:
    requests:
      cpu: 750m
      memory: 2Gi
    # limits:
    #   cpu: 2000m
    #   memory: 4Gi
  podAnnotations: {}
  nodeSelector: {}
  tolerations: []
  affinity: {}
  extraEnvs: [] # [{ name: AXIOM_XYZ, value: "xyz" }, ... ]

# Settings for axiomdb-query-fn
queryFn:
  replicas: 4
  resources:
    requests:
      cpu: 500m
      memory: 256Mi
    # limits:
    #   cpu: 2000m
    #   memory: 1Gi
  podAnnotations: {}
  nodeSelector: {}
  tolerations: []
  affinity: {}
  extraEnvs: [] # [{ name: AXIOM_XYZ, value: "xyz" }, ... ]

Configuring Queries on AWS

If you are deploying Axiom on AWS, we recommend running queries on Lambda, check out our guide on how to configure and run Queries on AWS Lambda

Install Axiom Using Helm

Once your values.yaml has been filled out, you can finally install Axiom in your Kubernetes cluster using the following commands:

helm upgrade -i -f values.yaml my-axiom axiom/axiom

Support