Skip to main content

GKE onboarding and best practices


  • Create a cluster in GKE Autopilot mode
    In     Autopilot mode, Google manages most of the infrastructure and provides a more managed K8s experience than GKE Standard mod

    Create an Autopilot cluster by specifying a name and region. After the cluster is created, you can deploy your workload through Kubernetes and Google will take care of the rest, including:

    •   Nodes: Automated node provisioning, scaling, and maintenance
    •   Networking: VPC-native traffic routing for public or private clusters
    •   Security: Shielded GKE Nodes and Workload Identity
    •   Telemetry: Cloud Operations logging and monitoring

  • When creating a cluster, Google cloud asks for Network configurations for the cluster
    Public clusters: Choose a public cluster to configure access from public networks to the cluster's workloads. Routes aren't created automatically. You cannot change this setting after the cluster is created.
    Private clusters: Choose a private cluster to assign internal IP addresses to Pods and nodes. This isolates the cluster's workloads from public networks. You cannot change this setting after the cluster is created. In this case, after you create the cluster, configure Cloud NAT to enable outbound internet connections from your cluster, Learn more 

  • Deployments  - a replicated, stateless application on your cluster

    Pods  - the smallest deployable unit in Kubernetes

    Services  - allow your application to receive traffic

    Autoscaling pods  - scale the application based on load or custom metrics

Comments

Post a Comment

Popular posts from this blog

Networking in Kubernetes

Pods that are running inside Kubernetes are running on a private, isolated network. By default they are visible from other pods and services within the same Kubernetes cluster, but not outside that network Every Pod has a unique IP address And it is reachable from all other Pods in the K8s cluster A pod is a host, just like your laptop, having an ip-address and a range of ports that can be alloted to containers A container runs on a specific port inside a pod In a Kubernetes environment, when services are deployed within the same namespace , they can communicate with each other using the service name as the hostname e.g. in the following snippet from appsetting.json form a .net core project, ' document-api'  is the name of the service "DocumentApiConfiguration" : { "BaseUrl" : "http://document-api/" } What if I want to access a service from another namespace? When you want to access a service from another namespace in Kubernetes, you typica...
Post a Comment
Read more

Terraform

It is an Infrastructure as Code tool Normally, if one has to configure VMs or other resources on the cloud, they have to go to the cloud provider's website and click a lot to get things done as supposed, terraform can do all of that provided you tell it precisely what to do in a .tf file e.g. which cloud provider you are using(GCP, Azure etc), which resource to configure with what specifications. One writes the file in hashicorp language (kinda like JSON) Free and Open source One has to install the CLI terraform init terraform apply  (to make the changes to cloud) terraform destroy
Post a Comment
Read more

Kubernetes

Some keywords: Node A Node is a worker machine in Kubernetes and may be either a virtual or a physical machine, depending on the cluster Clusters Kubernetes coordinates a highly available cluster of computers (nodes) that are connected to work as a single unit Namespace Namespaces are a way to organize clusters into virtual sub-clusters — they can be helpful when different teams or projects share a Kubernetes cluster . Any number of namespaces are supported within a cluster , each logically separated from others but with the ability to communicate with each other Kubernetes: Kubernetes is a portable, extesible open-source platform for managing and orchestration containerized workloads . It abstracts away complex container management tasks Provides us with declarative configuration to orchestrate containers in different computing environments This orchestration platform gives you the same ease of use and flexibility you might already know from Platform-as-a-Service (PaaS) or Infrastruct...
Post a Comment
Read more