This repository defines a Crossplane configuration package that demonstrates provisioning and using CAST AI.
The example platform supports the use case of
- ReadOnly Mode
- FullAccess Mode
The Infrastructure team deploys and manages the CAST AI infrastructure using the following Compositions:
flowchart LR
subgraph "Cluster Scoped"
direction LR
XR("XReadOnly")
MR1("Managed Resource \n(ekscluster.castai.upbound.io)")
MR2("Managed Resouce \n castai-agent \n(release.helm.crossplane.io)")
MR3("Secret \n name: castai-sample \n namespace: crossplane-system")
end
XR --> |"spec.resourceRef"| MR1 --> |"spec.writeConnectionSecretToRef"| MR3
XR --> |"spec.resourceRef"| MR2 --> |".valueFrom.secretKeyRef.name"| MR3
flowchart LR
subgraph "Cluster Scoped"
direction LR
XR("XFullAccess")
MR1("XR\n(xreadonly.aws.castai.io)")
MR2("Managed Resource \n(eksclusterid.castai.upbound.io)")
MR3("Managed Resouce \n castai-cluster-controller \n(release.helm.crossplane.io)")
MR4("Managed Resouce \n castai-evictor \n(release.helm.crossplane.io)")
MR5("Managed Resouce \n castai-spot-handler \n(release.helm.crossplane.io)")
MR6("Secret \n name: castai-sample \n namespace: crossplane-system")
MR7("Managed Resource \n(awsuserarn.castai.upbound.io)")
MR8("Managed Resource \n(nodeconfiguration.castai.upbound.io)")
MR9("Managed Resource \n(nodeconfigurationdefault.castai.upbound.io)")
MR10("Managed Resource \n AssumeRole \n(role.iam.aws.upbound.io)")
MR11("Managed Resource \n(rolepolicyattachment.iam.aws.upbound.io)")
MR12("Managed Resource \n(rolepolicyattachment.iam.aws.upbound.io)")
MR13("Managed Resource \n(policy.iam.aws.upbound.io)")
MR14("Managed Resource \n(rolepolicyattachment.iam.aws.upbound.io)")
MR15("Managed Resource \n(policy.iam.aws.upbound.io)")
MR16("Managed Resource \n(rolepolicyattachment.iam.aws.upbound.io)")
MR17("Managed Resource \n InstanceRole \n(role.iam.aws.upbound.io)")
MR18("Managed Resource \n(rolepolicyattachment.iam.aws.upbound.io)")
MR19("Managed Resource \n(rolepolicyattachment.iam.aws.upbound.io)")
MR20("Managed Resource \n(rolepolicyattachment.iam.aws.upbound.io)")
MR21("Managed Resource \n(instanceprofile.iam.aws.upbound.io)")
end
XR --> |"spec.resourceRef"| MR1
XR --> |"spec.resourceRef"| MR2 --> |"input"| MR7
XR --> |"spec.resourceRef"| MR3 --> |".valueFrom.secretKeyRef.name"| MR6
XR --> |"spec.resourceRef"| MR4 --> |".valueFrom.secretKeyRef.name"| MR6
XR --> |"spec.resourceRef"| MR5 --> |".valueFrom.secretKeyRef.name"| MR6
XR --> |"spec.resourceRef"| MR7 --> |"input"| MR10
XR --> |"spec.resourceRef"| MR8
XR --> |"spec.resourceRef"| MR9
XR --> |"spec.resourceRef"| MR10
XR --> |"spec.resourceRef"| MR11 --> |"selector"| MR10
XR --> |"spec.resourceRef"| MR12 --> |"selector"| MR10
XR --> |"spec.resourceRef"| MR13 --> |"selector"| MR14
XR --> |"spec.resourceRef"| MR14 --> |"selector"| MR10
XR --> |"spec.resourceRef"| MR15 --> |"selector"| MR16
XR --> |"spec.resourceRef"| MR16 --> |"selector"| MR10
XR --> |"spec.resourceRef"| MR17
XR --> |"spec.resourceRef"| MR18 --> |"selector"| MR17
XR --> |"spec.resourceRef"| MR19 --> |"selector"| MR17
XR --> |"spec.resourceRef"| MR20 --> |"selector"| MR17
XR --> |"spec.resourceRef"| MR21 --> |"selector"| MR17
First you will need access to a Kubernetes cluster. Ensure you are using the correct context:
kubectl config current-contextTo install Crossplane run:
helm repo add crossplane-stable https://charts.crossplane.io/stable
helm repo update
helm install crossplane \
--namespace crossplane-system \
--create-namespace crossplane-stable/crossplane Install the required Providers:
kubectl apply -f examples/providers.yamlYou can keep track of the provider install:
kubectl get -f examples/providers.yamlNext, install the CompositeResourceDefinitions and Compositions:
kubectl apply -f apis/XFullAccess -f apis/XReadOnlyThe Custom Platform APIs are Kubernetes CompositeResourceDefinition objects or XRD
for short. We can list them using kubectl:
kubectl get xrdThe following XRDs should be ESTABLISHED and OFFERED:
NAME ESTABLISHED OFFERED AGE
xfullaccess.aws.cast.ai True 5m
xreadonly.aws.cast.ai True 5mNow that Crossplane, the Provider and all the Compositions are installed we
need to give the provider AWS credentials. This is done by creating a ProviderConfig.
kubectl create secret generic aws-creds -n crossplane-system --from-file=creds=./creds.confWe will create the following ProviderConfig object to use the AWS credentials
from the previous step. See AUTHENTICATION for more authentication options like IRSA.
apiVersion: aws.upbound.io/v1beta1
kind: ProviderConfig
metadata:
name: default
spec:
credentials:
source: Secret
secretRef:
namespace: crossplane-system
name: aws-creds
key: credsNow that Crossplane, the Provider and all the Compositions are installed we
need to give the provider CAST AI credentials. This is done by creating a ProviderConfig.
kubectl create secret generic castai-creds -n crossplane-system --from-file=credentials=./castai.jsonWe will create the following ProviderConfig object to use the CAST AI credentials
apiVersion: castai.upbound.io/v1beta1
kind: ProviderConfig
metadata:
name: default
spec:
credentials:
source: Secret
secretRef:
namespace: crossplane-system
name: castai-creds
key: credentialsNow that Crossplane, the Provider and all the Compositions are installed we
need to give the provider helm RBAC access. This is done by creating a ProviderConfig.
Make sure provider-helm has enough permissions to install your chart into cluster We can give admin permissions by running:
SA=$(kubectl -n crossplane-system get sa -o name | grep provider-helm | sed -e 's|serviceaccount\/|crossplane-system:|g')
kubectl create clusterrolebinding provider-helm-admin-binding --clusterrole cluster-admin --serviceaccount="${SA}"We will create the following ProviderConfig object to use the CAST AI credentials
apiVersion: helm.crossplane.io/v1beta1
kind: ProviderConfig
metadata:
name: default
spec:
credentials:
source: InjectedIdentitykubectl apply -f examples/providerconfig-creds.yamlWe're now ready to deploy the examples.
Using files in the examples directory:
kubectl apply -f examples/XReadOnly.yaml
kubectl apply -f examples/XFullAccess.yamlApplying the examples to the cluster would create Kubernetes objects similar to the following:
kubectl get xreadonlyNAME SYNCED READY COMPOSITION AGE
sample-read-only True True xreadonly.aws.cast.ai 5skubectl get xfullaccessNAME SYNCED READY COMPOSITION AGE
sample True True xfullaccess.aws.cast.ai 5sTo Clean up the installation, run the following commands:
kubectl delete -f examples/XReadOnly.yaml -f examples/XFullAccess.yamlWait for all the cloud resources to be deleted:
kubectl get managedDelete the Compositions, Providers, and ProviderConfig after all the resources have been deleted.
kubectl delete -f apis/XReadOnly -f apis/XFullAccess
kubectl delete -f examples/providers.yaml
kubectl delete -f examples/providerconfig-creds.yamlkubectl delete secret -n crossplane-system aws-creds
kubectl delete secret -n crossplane-system castai-credsFor any questions, thoughts and comments don't hesitate to reach out or drop by slack.crossplane.io, and say hi to haarchri!
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent