Interacting with Kubernetes
You can deploy kdb Insights Enterprise to, AWS, GCP & Azure. The clusters are typically managed through Terraform GitOps scripts. New cluster setup and reconfigurations can be achieved by updating the terraform configuration and applying via DevOps processes.
Prerequisites
A number of tools are required to allow the user to deploy the kdb Insights Enterprise:
Connection to cluster
To connect to a cluster, you must create a context.
A context is the configuration that you use to access a particular cluster and namespace with a user account.
The user account must be authorized to access the cloud provider project. Use the cloud provider CLI to grant authorization. You may need assistance from the project administrator. See the relevant cloud provider documentation for further details on how to do this. Some info is provided below.
Assuming the user account has the appropriate permissions to the cloud provider and cluster, and has authenticated via the CLI, the next step is to connect to a specific cluster & namespace.
This command creates the necessary configuration to set up kubectl
to point to the cluster kxi-a1
in the region us-east-1
.
aws eks --region us-east-1 update-kubeconfig --name kxi-a1
To authorize gcloud
and utilize GKE, set the appropriate account and project to be active.
gcloud auth login
gcloud auth configure-docker
Note
The gcloud
command offers significant helpers. Running incomplete commands prompts you with available options. For example, gcloud container clusters
indicates cluster commands available. See gcloud
documentation for full details.
gcloud container clusters get-credentials \
kxi-g1 --region us-central1 --project <gcp-project-name>
To authorize az
and utilize AKS within the kubectl
setup the below steps are required.
- Set the account subscription id to link
az
to the Azure subscription.
az account show | grep id # Get subscription ID
az account set --subscription <id> # Set subscription ID
- Link
kubectl
to your credentials by providing the resource group name containing the AKS cluster along with the cluster name.
az aks get-credentials --resource-group <resource group name> --name <cluster name>
Kubernetes contexts and namespaces
Once connected to the cluster, set up an 'active context'. The default context points to the default namespace within the active cluster. Change this to a siloed namespace to allow multiple deployments of applications within a cluster.
Depending on the maturity of the cluster being used and the kubectl
setup, you may need to carry out a number of steps. See full documentation on contexts and namespaces. Below are common commands to help interrogate the active context, clusters, and namespaces.
View active context
Displays the name of the active context.
kubectl config current-context
View defined contexts
Displays all contexts within your configuration and includes cluster information, users access configuration and namespace association. Note this is dynamically built from various calls highlighted here.
kubectl config view
Create from command line
kubectl create namespace kxi-sys
Create a context
Link the cluster, user and namespace to a single context entity.
kubectl config set-context kxi --namespace=kxi-sys \
--cluster=<cluster-name> \
--user=<user-name>
Note
The namespace, cluster and user should already exist in your config kubectl config view
.
Set active context
Once created, set the context as active.
kubectl config use-context kxi