Open WebUI Helm

Chart to deploy Open WebUI, a ChatGPT-Style Web UI Client for Ollama 🦙.

TL;DR

helm repo add braveokafor https://braveokafor.github.io/helm-charts/
helm install ollama braveokafor/open-webui

OR:

helm repo add braveokafor https://charts.braveokafor.com
helm install ollama braveokafor/open-webui

Introduction

Deploy Open WebUI on Kubernetes easily with this Helm chart.

It enables you to run a ChatGPT-style web UI client, with a variety of open-source large language models available at ollama.ai/library.

Alternatively, you can also also run against the OpenAI API with an API Key.

Source Code

• https://github.com/jmorganca/ollama/
• https://github.com/open-webui/open-webui
• https://github.com/braveokafor/open-webui-helm/

Requirements

Repository	Name	Version
https://charts.braveokafor.com	ollama	0.1.x

Maintainers

Name	Email	Url
braveokafor	[email protected]	https://braveokafor.com/

Installing the Chart

To install the chart with the release name ollama:

helm install ollama braveokafor/open-webui

The command deploys Ollama on the Kubernetes cluster in the default configuration.

The Parameters section lists the parameters that can be configured during installation.

Uninstalling the Chart

To uninstall/delete the ollama deployment:

helm delete ollama

The command removes all the Kubernetes components associated with the chart and deletes the release.

Configuration and installation details

Specify each parameter using the --set key=value[,key=value] argument to helm install. For example,

helm install ollama \
  --set ollama.gpu.enabled=true \
    braveokafor/open-webui

The above command enables GPU support for Ollama.

Alternatively, a YAML file that specifies the values for the above parameters can be provided while installing the chart. For example:

helm install ollama -f values.yaml braveokafor/open-webui

Example fully configured values.yaml to setup with Ollama:

# values.yaml
ollama:
  install: true
  enabled: true
  ollama:
    gpu:
      enabled: true
    models:
      - gemma
      - mistral
  nodeSelector:
    cloud.google.com/gke-accelerator: "nvidia-tesla-t4"
resources:
  requests:
    memory: 2048Mi
    cpu: 1000m
ingress:
  enabled: true
  hostname: ollama.braveokafor.com
  annotations:
    kubernetes.io/ingress.global-static-ip-name: ollama
    ingressClassName: gce
  tls:
    enabled: true

Example fully configured values.yaml to setup with OpenAI:

# values.yaml
openai:
  enabled: true
  apiKey: "DX0mvZ6hQvUpM3FSL9qj58KTsbFwPYNsvgT0ztAiYA2q-eWsrxoQ"
ollama:
  install: false
  enabled: false
resources:
  requests:
    memory: 2048Mi
    cpu: 1000m
ingress:
  enabled: true
  hostname: ollama.braveokafor.com
  annotations:
    kubernetes.io/ingress.global-static-ip-name: ollama
    ingressClassName: gce
  tls:
    enabled: true

Overrides

Key	Type	Default	Description
fullnameOverride	string	""	String to fully override "webui"
nameOverride	string	"webui"	Provide a name in place of webui

Ollama parameters

Key	Type	Default	Description
ollama.enabled	bool	true	Inject Ollama API Environment Variables
ollama.externalURL	string	""	External Ollama API URL (if Chart is not installed)
ollama.install	bool	true	Install Ollama Helm Chart

OpenAI parameters

Key	Type	Default	Description
openai.apiKey	string	""	OpenAI API Key
openai.baseUrl	string	"https://api.openai.com/v1"	OpenAI API API URL
openai.enabled	bool	false	Inject OpenAI API Environment Variables
openai.existingApiKeySecret	string	""	Name of an existing secret resource containing the OpenAI API credentials
openai.existingApiKeySecretKey	string	"openai-api-key"	Name of the existing secret's key containing the OpenAI API credentials

Image parameters

Key	Type	Default	Description
image.imagePullPolicy	string	"IfNotPresent"	Image pull policy for Open WebUI
image.repository	string	"ghcr.io/open-webui/open-webui"	Repository to use for Open WebUI
image.tag	string	"v0.1.106"	Tag to use for Open WebUI
imagePullSecrets	list	[]	Secrets with credentials to pull images from a private registry

General parameters

Key	Type	Default	Description
affinity	object	{}	Assign custom [affinity] rules to the deployment
autoscaling.behavior	object	{}	Configures the scaling behavior of the target in both Up and Down directions.
autoscaling.enabled	bool	false	Enable Horizontal Pod Autoscaler ([HPA]) for Open WebUI
autoscaling.maxReplicas	int	5	Maximum number of replicas for Open WebUI [HPA]
autoscaling.metrics	list	[]	Configures custom HPA metrics for Open WebUI Ref: https://kubernetes.io/docs/tasks/run-application/horizontal-pod-autoscale/
autoscaling.minReplicas	int	1	Minimum number of replicas for Open WebUI [HPA]
autoscaling.targetCPUUtilizationPercentage	int	50	Average CPU utilization percentage for Open WebUI [HPA]
autoscaling.targetMemoryUtilizationPercentage	int	50	Average memory utilization percentage for Open WebUI [HPA]
containerPorts.http	int	8080	webui container port
containerSecurityContext	object	See [values.yaml]	Open WebUI container-level security context
deploymentAnnotations	object	{}	Annotations to be added to webui Deployment
deploymentStrategy	object	{}	Deployment strategy to be added to the webui Deployment
dnsConfig	object	{}	[DNS configuration]
dnsPolicy	string	"ClusterFirst"	Alternative DNS policy for Open WebUI pods
env	list	[]	Environment variables to pass to Open WebUI
envFrom	list	[] (See [values.yaml])	envFrom to pass to Open WebUI
extraArgs	list	[]	Additional command line arguments to pass to Open WebUI
hostAliases	list	[]	Mapping between IP and hostnames that will be injected as entries in the pod's hosts files
hostNetwork	bool	false	Host Network for Open WebUI pods
lifecycle	object	{}	Specify postStart and preStop lifecycle hooks for your webui container Specifying this will disable pulling models on container start-up
livenessProbe.failureThreshold	int	3	Minimum consecutive failures for the [probe] to be considered failed after having succeeded
livenessProbe.initialDelaySeconds	int	10	Number of seconds after the container has started before [probe] is initiated
livenessProbe.periodSeconds	int	10	How often (in seconds) to perform the [probe]
livenessProbe.successThreshold	int	1	Minimum consecutive successes for the [probe] to be considered successful after having failed
livenessProbe.timeoutSeconds	int	1	Number of seconds after which the [probe] times out
nodeSelector	object	{}	[Node selector]
pdb.annotations	object	{}	Annotations to be added to Open WebUI pdb
pdb.enabled	bool	false	Deploy a [PodDisruptionBudget] for Open WebUI
pdb.labels	object	{}	Labels to be added to Open WebUI pdb
pdb.maxUnavailable	string	""	Number of pods that are unavailable after eviction as number or percentage (eg.: 50%).
pdb.minAvailable	string	"" (defaults to 0 if not specified)	Number of pods that are available after eviction as number or percentage (eg.: 50%)
podAnnotations	object	{}	Annotations to be added to webui pods
podLabels	object	{}	Labels to be added to webui pods
priorityClassName	string	""	Priority class for Open WebUI pods
readinessProbe.failureThreshold	int	3	Minimum consecutive failures for the [probe] to be considered failed after having succeeded
readinessProbe.initialDelaySeconds	int	10	Number of seconds after the container has started before [probe] is initiated
readinessProbe.periodSeconds	int	10	How often (in seconds) to perform the [probe]
readinessProbe.successThreshold	int	1	Minimum consecutive successes for the [probe] to be considered successful after having failed
readinessProbe.timeoutSeconds	int	1	Number of seconds after which the [probe] times out
replicas	int	1	The number of webui pods to run
resources	object	{}	Resource limits and requests for Open WebUI
revisionHistoryLimit	int	3	Number of old deployment ReplicaSets to retain. The rest will be garbage collected.
securityContext	object	{} (See [values.yaml])	Toggle and define pod-level security context.
terminationGracePeriodSeconds	int	30	terminationGracePeriodSeconds for container lifecycle hook
tolerations	list	[]	[Tolerations] for use with node taints
topologySpreadConstraints	list	[]	Assign custom [TopologySpreadConstraints] rules to Open WebUI # Ref: https://kubernetes.io/docs/concepts/workloads/pods/pod-topology-spread-constraints/ # If labelSelector is left out, it will default to the labelSelector configuration of the deployment
volumeMounts	list	[]	Additional volumeMounts to the webui container
volumes	list	[]	Additional volumes to the webui pod

Service parameters

Key	Type	Default	Description
service.annotations	object	{}	webui service annotations
service.externalIPs	list	[]	webui service external IPs
service.externalTrafficPolicy	string	""	Denotes if this Service desires to route external traffic to node-local or cluster-wide endpoints
service.labels	object	{}	webui service labels
service.loadBalancerIP	string	""	LoadBalancer will get created with the IP specified in this field
service.loadBalancerSourceRanges	list	[]	Source IP ranges to allow access to service from
service.nodePortHttp	int	30080	webui service http port for NodePort service type (only if webui.service.type is set to "NodePort")
service.servicePortHttp	int	80	webui service http port
service.servicePortHttpName	string	"http"	webui service http port name, can be used to route traffic via istio
service.sessionAffinity	string	""	Used to maintain session affinity. Supports ClientIP and None
service.type	string	"ClusterIP"	webui service type
serviceAccount.annotations	object	{}	Annotations applied to created service account
serviceAccount.automountServiceAccountToken	bool	true	Automount API credentials for the Service Account
serviceAccount.create	bool	true	Create webui service account
serviceAccount.labels	object	{}	Labels applied to created service account
serviceAccount.name	string	"webui"	webui service account name

Persistence parameters

Key	Type	Default	Description
persistence.accessModes	list	["ReadWriteOnce"]	Persistent Volume Access Modes
persistence.annotations	object	{}	Persistent Volume Claim annotations
persistence.dataSource	object	{}	Custom PVC data source
persistence.enabled	bool	true	Create a PVC for Open WebUI (empty-dir will be used if not specified)
persistence.existingClaim	string	""	The name of an existing PVC to use for persistence
persistence.selector	object	{}	Selector to match an existing Persistent Volume for Open WebUI data PVC # If set, the PVC can't have a PV dynamically provisioned for it
persistence.size	string	"4Gi"	Size of data volume
persistence.storageClass	string	""	Storage class of backing PVC

Service account parameters

Key	Type	Default	Description
serviceAccount.annotations	object	{}	Annotations applied to created service account
serviceAccount.automountServiceAccountToken	bool	true	Automount API credentials for the Service Account
serviceAccount.create	bool	true	Create webui service account
serviceAccount.labels	object	{}	Labels applied to created service account
serviceAccount.name	string	"webui"	webui service account name

Ingress parameters

Key	Type	Default	Description
ingress.annotations	object	{}	Additional ingress annotations
ingress.enabled	bool	false	Enable an ingress resource for Open WebUI
ingress.extraHosts	list	[] (See [values.yaml])	The list of additional hostnames to be covered by ingress record
ingress.extraPaths	list	[] (See [values.yaml])	Additional ingress paths
ingress.extraRules	list	[] (See [values.yaml])	Additional ingress rules
ingress.extraTls	list	[] (See [values.yaml])	Additional TLS configuration
ingress.hostname	string	""	Open WebUI hostname
ingress.ingressClassName	string	""	Defines which ingress controller will implement the resource
ingress.labels	object	{}	Additional ingress labels
ingress.path	string	"/"	The path to Open WebUI
ingress.pathType	string	"Prefix"	Ingress path type. One of Exact, Prefix or ImplementationSpecific
ingress.tls.enabled	bool	false	Enable TLS configuration for the hostname defined at ingress.hostname
ingress.tls.selfSigned	bool	false	Create a Self Signed TLS Certificate for Open WebUI

---

Autogenerated from chart metadata using helm-docs