Arrow Flight values#

Values under arrowflight configure the Arrow Flight server, which serves fast reads of feature groups and training datasets.

Generated from the Hopsworks Helm chart 5.2.0-alpha-1791549041 (Hopsworks 5.2.0).

Deployed when global._hopsworks.full_platform is true.

General#

Defaults as YAML

arrowflight:
  appName: arrowflight
  common:
    monitoring_port: 12810
    port: 5005
  configmap:
    name: arrowflight-configmap
  hopsworkslib: {}
  hpa:
    enabled: true
    flyingduck_queue_time_avg: 27
    flyingduck_request_queue_gauge: 2
    maxReplicas: 3
  image:
    pullPolicy: IfNotPresent
    registry: docker.hops.works
  nodeSelector: {}
  podDisruptionBudget:
    enabled: true
    minAvailable: 1
  spillVolume:
    size: 20Gi
    storageClassName: null
  tolerations: []
  topologySpreadConstraint: {}

arrowflight #: Type object, default {}. override arrow flight values
arrowflight.appName #: Type string, default "arrowflight".
arrowflight.common.monitoring_port #: Type int, default 12810.
arrowflight.common.port #: Type int, default 5005.
arrowflight.configmap.name #: Type string, default "arrowflight-configmap".
arrowflight.hopsworkslib #: Type object, default {}. override hopsworkslib values
arrowflight.hpa.enabled #: Type bool, default true. Autoscale the Deployment with an HPA (suppressed when a VPA targets it). While active, deployment.replicas is not rendered and the HPA owns spec.replicas.
arrowflight.hpa.flyingduck_queue_time_avg #: Type int, default 27.
arrowflight.hpa.flyingduck_request_queue_gauge #: Type int, default 2.
arrowflight.hpa.maxReplicas #: Type int, default 3.
arrowflight.image #: Type object, default {"pullPolicy":"IfNotPresent","registry":"docker.hops.works"}. image configuration. The image tag is the .Chart.AppVersion
arrowflight.nodeSelector #: Type object, default {}. node selector configuration
arrowflight.podDisruptionBudget.enabled #: Type bool, default true.
arrowflight.podDisruptionBudget.minAvailable #: Type int, default 1.
arrowflight.spillVolume #: Type object, default {"size":"20Gi","storageClassName":null}. storage configuration for the ephemeral volume used by DuckDB for spilling on disk during query execution for spilling on disk during query execution for spilling on disk during query execution
arrowflight.spillVolume.size #: Type string, default "20Gi". size of the ephemeral volume
arrowflight.spillVolume.storageClassName #: Type string, default nil. storage class name. If null, the default storage class will be used
arrowflight.tolerations #: Type list, default [].
arrowflight.topologySpreadConstraint #: Type object, default {}. The default topology spread constraint. If not defined the global topology spread constraint would be used instead.

deployment#

Defaults as YAML

arrowflight:
  deployment:
    annotations:
      prometheus.io/path: /metrics
      prometheus.io/port: '12810'
      prometheus.io/scheme: http
      prometheus.io/scrape: 'true'
    name: arrowflight-deployment
    replicas: 1
    resources:
      limits:
        memory: 8192Mi
      requests:
        cpu: '2'
        memory: 6553Mi
    security:
      runAsGroup: 1520
      runAsUser: 1525
    server:
      hopsfs_query_mode: view
      memory_limit: '6'
      queue_timeout: '600'

arrowflight.deployment.annotations."prometheus.io/path" #: Type string, default "/metrics".
arrowflight.deployment.annotations."prometheus.io/port" #: Type string, default "12810".
arrowflight.deployment.annotations."prometheus.io/scheme" #: Type string, default "http".
arrowflight.deployment.annotations."prometheus.io/scrape" #: Type string, default "true".
arrowflight.deployment.name #: Type string, default "arrowflight-deployment".
arrowflight.deployment.replicas #: Type int, default 1. Number of replicas. Not rendered while the HPA is active (hpa.enabled and no VPA on this Deployment): the HPA then owns spec.replicas and this value is its minReplicas. Turning the HPA on for a running release drops the Deployment to 1 once, until the HPA scales it back up.
arrowflight.deployment.resources.limits #: Type object, default {"memory":"8192Mi"}. resources limits configuration
arrowflight.deployment.resources.requests #: Type object, default {"cpu":"2","memory":"6553Mi"}. resources requests configuration
arrowflight.deployment.security.runAsGroup #: Type int, default 1520.
arrowflight.deployment.security.runAsUser #: Type int, default 1525.
arrowflight.deployment.server.hopsfs_query_mode #: Type string, default "view".
arrowflight.deployment.server.memory_limit #: Type string, default "6".
arrowflight.deployment.server.queue_timeout #: Type string, default "600".

externalLoadBalancer#

Defaults as YAML

arrowflight:
  externalLoadBalancer:
    annotations: {}
    class: null
    enabled: null
    managed: null
    nodePort: null
    nodeSelector: {}

arrowflight.externalLoadBalancer.annotations #: Type object, default {}. annotations for load balancer
arrowflight.externalLoadBalancer.class #: Type string, default nil. load balancer class name
arrowflight.externalLoadBalancer.enabled #: Type string, default nil. Enable External Load Balancers for Arrowflight server. If not set the .global._hopsworks.externalLoadBalancers.enabled will be used instead
arrowflight.externalLoadBalancer.managed #: Type string, default nil. Cloud provider provisions Load Balancers. If not set the .global._hopsworks.externalLoadBalancers.managed will be used instead
arrowflight.externalLoadBalancer.nodePort #: Type string, default nil. Explicit nodePort for the external service when the load balancer is unmanaged (managed: false), so a load balancer outside Kubernetes can target a fixed port. Null lets Kubernetes allocate one from the cluster's node-port range; a set value must lie in that range (30000-32767 by default), which the API server enforces at install.
arrowflight.externalLoadBalancer.nodeSelector #: Type object, default {}. selector for nodes the load balancer can use to route traffic

service#

Defaults as YAML

arrowflight:
  service:
    annotations:
      consul.hashicorp.com/service-name: flyingduck
      consul.hashicorp.com/service-port: server
      consul.hashicorp.com/service-tags: server
      prometheus.io/path: /metrics
      prometheus.io/port: '12810'
      prometheus.io/scheme: http
      prometheus.io/scrape: 'true'
    name: arrowflight-server

arrowflight.service.annotations."consul.hashicorp.com/service-name" #: Type string, default "flyingduck".
arrowflight.service.annotations."consul.hashicorp.com/service-port" #: Type string, default "server".
arrowflight.service.annotations."consul.hashicorp.com/service-tags" #: Type string, default "server".
arrowflight.service.annotations."prometheus.io/path" #: Type string, default "/metrics".
arrowflight.service.annotations."prometheus.io/port" #: Type string, default "12810".
arrowflight.service.annotations."prometheus.io/scheme" #: Type string, default "http".
arrowflight.service.annotations."prometheus.io/scrape" #: Type string, default "true".
arrowflight.service.name #: Type string, default "arrowflight-server".