Service Catalog is a Kubernetes Incubator project that provides a Kubernetes-native workflow for integrating with Open Service Brokers to provision and bind to application dependencies like databases, object storage, message-oriented middleware, and more.
For more information, visit the project on github.
- Kubernetes 1.13+
charts/catalog
already exists in your local machine
To install the chart with the release name catalog
:
$ helm install . --name catalog --namespace catalog
To uninstall/delete the catalog
deployment:
$ helm delete --purge catalog
The command removes all the Kubernetes components associated with the chart and deletes the release.
The following tables lists the configurable parameters of the Service Catalog chart and their default values.
Parameter | Description | Default |
---|---|---|
image |
Service catalog image to use | quay.io/kubernetes-service-catalog/service-catalog:v0.3.1 |
imagePullPolicy |
imagePullPolicy for the service catalog |
Always |
imagePullSecrets |
The pre-existing secrets to use to pull images from a private registry | [] |
webhook.updateStrategy |
updateStrategy for the service catalog webhook deployment |
RollingUpdate |
webhook.minReadySeconds |
how many seconds an webhook server pod needs to be ready before killing the next, during update | 1 |
webhook.annotations |
Annotations for webhook pods | {} |
webhook.nodeSelector |
A nodeSelector value to apply to the webhook pods. If not specified, no nodeSelector will be applied | |
webhook.service.type |
Type of service; valid values are LoadBalancer , NodePort and ClusterIP |
ClusterIP |
webhook.service.nodePort.securePort |
If service type is NodePort , specifies a port in allowable range (e.g. 30000 - 32767 on minikube); The TLS-enabled endpoint will be exposed here |
30443 |
webhook.service.clusterIP |
If service type is ClusterIP, specify clusterIP as None for headless services OR specify your own specific IP OR leave blank to let Kubernetes assign a cluster IP |
|
webhook.verbosity |
Log level; valid values are in the range 0 - 10 | 10 |
webhook.healthcheck.enabled |
Enable readiness and liveliness probes | true |
webhook.resources |
Resources allocation (Requests and Limits) | {requests: {cpu: 100m, memory: 20Mi}, limits: {cpu: 100m, memory: 30Mi}} |
controllerManager.replicas |
replicas for the service catalog controllerManager pod count |
1 |
controllerManager.updateStrategy |
updateStrategy for the service catalog controllerManager deployments |
RollingUpdate |
controllerManager.minReadySeconds |
how many seconds a controllerManager pod needs to be ready before killing the next, during update | 1 |
controllerManager.annotations |
Annotations for controllerManager pods | {} |
controllerManager.nodeSelector |
A nodeSelector value to apply to the controllerManager pods. If not specified, no nodeSelector will be applied | |
controllerManager.healthcheck.enabled |
Enable readiness and liveliness probes | true |
controllerManager.verbosity |
Log level; valid values are in the range 0 - 10 | 10 |
controllerManager.resyncInterval |
How often the controller should resync informers; duration format (20m , 1h , etc) |
5m |
controllerManager.osbApiRequestTimeout |
The maximum amount of timeout to any request to the broker; duration format (60s , 3m , etc) |
60s |
controllerManager.brokerRelistInterval |
How often the controller should relist the catalogs of ready brokers; duration format (20m , 1h , etc) |
24h |
controllerManager.brokerRelistIntervalActivated |
Whether or not the controller supports a --broker-relist-interval flag. If this is set to true, brokerRelistInterval will be used as the value for that flag. | true |
controllerManager.profiling.disabled |
Disable profiling via web interface host:port/debug/pprof/ | false |
controllerManager.profiling.contentionProfiling |
Enables lock contention profiling, if profiling is enabled | false |
controllerManager.leaderElection.activated |
Whether the controller has leader election enabled | false |
controllerManager.serviceAccount |
Service account | service-catalog-controller-manager |
controllerManager.enablePrometheusScrape |
Whether the controller will expose metrics on /metrics | false |
controllerManager.resources |
Resources allocation (Requests and Limits) | {requests: {cpu: 100m, memory: 20Mi}, limits: {cpu: 100m, memory: 30Mi}} |
controllerManager.service.type |
Type of service; valid values are LoadBalancer , NodePort and ClusterIP |
ClusterIP |
controllerManager.service.nodePort.securePort |
If service type is NodePort , specifies a port in allowable range (e.g. 30000 - 32767 on minikube); The TLS-enabled endpoint will be exposed here |
30444 |
controllerManager.service.clusterIP |
If service type is ClusterIP, specify clusterIP as None for headless services OR specify your own specific IP OR leave blank to let Kubernetes assign a cluster IP |
|
rbacEnable |
If true, create & use RBAC resources | true |
originatingIdentityEnabled |
Whether the OriginatingIdentity feature should be enabled | true |
persistence.storageClass |
Define the storageclass use by pvc | null |
affinity |
Affinity settings (docs) | {} |
asyncBindingOperationsEnabled |
Whether or not alpha support for async binding operations is enabled | false |
namespacedServiceBrokerDisabled |
Whether or not alpha support for namespace scoped brokers is disabled | false |
nodeSelector |
Node labels for pod assignment (global parameter for all pods) | {} |
podLabels |
Additional pod labels to include for all pods | {} |
priorityClassName |
Define PriorityClass for pods | "" |
tolerations |
Tolerations for pod assignment | [] |
Specify each parameter using the --set key=value[,key=value]
argument to
helm install
.
Alternatively, a YAML file that specifies the values for the parameters can be provided while installing the chart. For example:
$ helm install . --name catalog --namespace catalog --values values.yaml