Create a Kubernetes cluster w/ vSphere

Quick overview on how to create VMware vSphere Kubernetes clusters through the Kublr Platform.

Deploy simple 1 master and 2 work node Kubernetes clusters for demo purposes.

First download and install Kublr-in-a-box at https://kublr.com/demo.

System requirements for the cluster nodes

x86 64-bit hardware
At least 5Gb of RAM and 2 vCPU per master node
At least 16Gb og RAM and 8 vCPU in sum for work nodes if you plan deploying Kublr Platform
Supported OS on nodes: Ubuntu 16.04 LTS and above, RHEL 7.4 and above, CentOS 7.4 and above
Direct connection to the virtual machine network
Direct connection to vCenter and ESXi hosts
100Mb stable internet connection

Kublr can use two kinds of VM templates:

With installed vm-tools (This case is used for custom VM templates)
With installed cloud-init (This case is used for prepared VM templates from the vendor)

For a quick start, you can use a prepared VM template with installed cloud-init from ova template.

You can upload this template to vCenter in two ways:

Click Clusters from the menu in the left navigation menu bar
Click the Add Cluster
Enter a cluster name
Select the provider VMware vSphere
Create VMware vSphere Credentials or select existing credentials
Data center will be loaded according your …
Select Resource Pool
Note: Before creating a cluster, make sure that sufficient resources are reserved in the resource pool. If the resource pool type is expandable, then it can borrow resources from the parent resource pool. If resource pool type is fixed and it does not has enough resources then cluster creation cannot be performed. More information about resource pool you can find in your vCenter
Select Datastore Type and Datastore
Select vSphere Network that will be used by cluster. Kublr supports all network types provided by vSphere including NSX-T
Select Number of Master Nodes and VM Template.
For some Templates you should create Username/Password Credentials used by Kublr to access VM during initialization
Select the number of work nodes. In the advanced options you should set RAM at more than the default (e.g. 6000 MB)
Click Confirm and Install button
Wait until cluster is switched to Running status