Skip to content

Latest commit

 

History

History
86 lines (66 loc) · 3.1 KB

installation-gcp.md

File metadata and controls

86 lines (66 loc) · 3.1 KB
Raw

Google Cloud Platform

In here, we'll show you how to setup a kubernetes cluster in GCP.

Kubernetes Engine

In order to create a kubernetes cluster, follow the below steps:

  1. Create a New project in you dashboard.

  2. Choose Kubernetes Engine from Compute section in the gcp dashboard. Google Cloud Platform Dashboard

  3. You have to go to the this page: Kubernete Engine

  4. In order to create a new cluster, push Create cluster botton. You can choose the version of kubernetes which you want to use and Zone of your cluster. Create Kubernetes Cluster

  5. You can choose the number of workers in NODE POOLS section: select default-pool Default Pool

  6. Select the Nodes section from default-pool. (If you want to customize the specifications (e.g. RAM and vCPU) of nodes do this)
    Node Specifications

  7. Push Create button and wait for creating cluster. After a few minutes you can see something like this Cluster

  8. You can see the details of your cluster by clicking the name of your cluster. Cluster Details

Your Kubernetes Cluster is ready now.

Connect to the Cluster

We want to connect to the cluster from our computer, so we need to configure gcloud SDK in our system.

Now it needs to configure the kubectl to access to our cluster. We can get the credentials of our cluster by pushing the Connect botton in dashboard gcloud Connect copy the command-line access and run it into you terminal to create the config file of your cluster in ~/.kube/config path.

Opening sdghafouri/utilization-server-smart-scheduler Port in the firewall (gcloud)

ّImportant step: We need to open to access to port of sdghafouri/utilization-server-smart-scheduler from outside the cluster:

UTILIZATION_PORT=30000
gcloud compute firewall-rules create sdghafouri/utilization-server-smart-scheduler-node-port --allow tcp:${UTILIZATION_PORT}

kubectl

In order to see you nodes, run the following command in your system:

kubectl get nodes
  • you can see something like this: kubectl get nodes

we want to run a pod in a specific node, so write the following content into file with name nginx_node_1.yaml:

apiVersion: v1
kind: Pod
metadata:
  name: nginx-1
  labels:
    web: nginx
spec:
  containers:
  - name: nginx-1
    image: nginx
    imagePullPolicy: IfNotPresent
  nodeName: gke-kubernetes-cosolidat-default-pool-1041ab9e-cdt1

Run the command below to start a pod:

kubectl apply -f ./nginx_node_1.yaml
  • run kubectl get pods -o wide and then you can see something like this: kubectl get pods