Kubernetes provides excellent container orchestration, but setting up a Kubernetes cluster from scratch can be painful. One solution is to use Kubernetes Operations, or kOps.
What is kOps?
kOps, also known as Kubernetes operations, is an open-source project which helps you create, destroy, upgrade, and maintain a highly available, production-grade Kubernetes cluster. Depending on the requirement, kOps can also provision cloud infrastructure.
kOps is mostly used in deploying AWS and GCE Kubernetes clusters. But officially, the tool only supports AWS. Support for other cloud providers (such as DigitalOcean, GCP, and OpenStack) are in the beta stage at the time of writing this article. The developers of this open source project describe it as “kubectl for clusters” on their Github homepage which may be confusing for some but helpful for others. Let’s start by reviewing its benefits.
Benefits of kOps
Automates the provisioning of AWS and GCE Kubernetes clusters
Deploys highly available Kubernetes masters
Supports rolling cluster updates
Autocompletion of command in the command line
Generates Terraform and CloudFormation configurations
Manages cluster add-ons
Supports state-sync model for dry-runs and automatic idempotency
Creates instance groups to support heterogeneous clusters
ubuntu@ip-161-32-38-140:~$ kOps
kOps is Kubernetes ops.
kOps is the easiest way to get a production grade Kubernetes cluster up and running. We like to think of it as kubectl for clusters.
kOps helps you create, destroy, upgrade and maintain production-grade, highly available, Kubernetes clusters from the command line. AWS (Amazon Web Services) is currently officially supported, with GCE and VMware vSphere in alpha support.
Usage:
kOps [command]
Available Commands:
completion Output shell completion code for the given shell (bash or zsh).
create Create a resource by command line, filename or stdin.
delete Delete clusters,instancegroups, or secrets.
describe Describe a resource.
edit Edit clusters and other resources.
export Export configuration.
get Get one or many resources.
import Import a cluster.
replace Replace cluster resources.
rolling-update Rolling update a cluster.
toolbox Misc infrequently used commands.
update Update a cluster.
upgrade Upgrade a kubernetes cluster.
validate Validate a kOps cluster.
version Print the kOps version information.
Flags:
--alsologtostderr log to standard error as well as files
--config string config file (default is $HOME/.kOps.yaml)
-h, --help help for kOps
--log_backtrace_at traceLocation when logging hits line file:N, emit a stack trace (default :0)
--log_dir string If non-empty, write log files in this directory
--logtostderr log to standard error instead of files (default false)
--name string Name of cluster
--state string Location of state storage
--stderrthreshold severity logs at or above this threshold go to stderr (default 2)
-v, --v Level log level for V logs
--vmodule moduleSpec comma-separated list of pattern=N settings for file-filtered logging
Use "kOps [command] --help" for more information about a command.
Check the kOps version.
ubuntu@ip-161-32-38-140:~$ kOps version
Version 1.8.0 (git-5099bc5)
Pick the ideal instance type for your workload using an ML-powered visual catalog map
Below are the most widely-used kOps commands you should know.
kOps create
This kOps command is used to register a cluster:
kOps create cluster <clustername>
There are many other parameters which you can add in addition to the default command, like zone, region, instance type, and number of nodes.
kOps update
This kOps command is used to update the cluster with the specified cluster specification:
kOps update cluster --name <clustername>
We recommend running this command in preview mode when you need to triangulate your desired output. Run the command with the --yes flag to apply your changes to the cluster.
kOps get
This kOps command is used to list all the clusters.
kOps get clusters
kOps delete
This command deletes a specified cluster from the registry along with all of its cloud resources.
kOps delete cluster --name <clustername>
You can also run this command in preview mode, just like update.
Identify under/over-provisioned K8s resources and use Terraform to auto-optimize
This kOps command is used to update a Kubernetes cluster to match the cloud and kOps specifications.
kOps rolling-update cluster --name <clustername>
You can run this command also in the preview mode just like update.
kOps validate
This command validates if the created cluster is up or not. If the nodes and pods are still in the pending state, the validate command returns that the cluster is not healthy yet.
kOps validate cluster --wait <specified_time>
This command waits and validates the cluster for 10 minutes.
Now that you know the basics of kOps commands, let’s look at how to create a Kubernetes cluster on AWS using kOps.
How to Set Up Kubernetes on AWS using kOps
Install kubectl.
ubuntu@ip-161-32-38-140:~$ curl -LO https://storage.googleapis.com/kubernetes-release/release/$(curl -s https://storage.googleapis.com/kubernetes-release/release/stable.txt)/bin/linux/amd64/kubectl
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 38.3M 100 38.3M 0 0 6862k 0 0:00:05 0:00:05 --:--:-- 8127k
Give the executable permission to the downloaded file and move it to /usr/local/bin/
Create an S3 Bucket using the AWS CLI where kOps will save all the cluster's state information. Use LocationConstraint to avoid any error with the region.
Generate ssh key for which will be used by kOps for cluster login and password generation.
ubuntu@ip-161-32-38-140:~$ ssh-keygen
Generating public/private rsa key pair.
Enter file in which to save the key (/home/ubuntu/.ssh/id_rsa):
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /home/ubuntu/.ssh/id_rsa.
Your public key has been saved in /home/ubuntu/.ssh/id_rsa.pub.
The key fingerprint is:
SHA256:fH4JCBXMNRqzk1hmoK+cXmwSFaeBsuGBA5IWMkNuvq0 ubuntu@ip-161-32-38-140
The key's randomart image is:
+---[RSA 2048]----+
|O=. .++Xoo |
|B++ .. @o* . |
|.= =. = = |
|o o o o o |
| . . . S o |
| o. = o . . |
| . .= + . o |
| .. + . |
| E . |
+----[SHA256]-----+
Expose the cluster name and s3 bucket as environment variables. This is only applicable for the current session. The suffix ‘.k8s.local’ is used because we are not using any pre-configured DNS.
Use the kOps create command to create the cluster. Below are the parameters we used:
--cloud
tells which cloud provider I am using
--zones
is the zone where the cluster instance will get deployed
--node-count
is the number of nodes I want in the Kubernetes cluster I am deploying
--node-size / --master-size
are the ec2 instance types, I am using the micro instances
--name
is the name of the cluster
ubuntu@ip-161-32-38-140:~$ kOps create cluster --cloud=aws --zones=eu-central-1a --node-count=1 --node-size=t2.micro --master-size=t2.micro --name=${kOps_CLUSTER_NAME}
I0216 17:14:49.225238 3547 subnets.go:180] Assigned CIDR 172.20.32.0/19 to subnet eu-central-1a
I0216 17:14:50.068088 3547 create_cluster.go:717] Using SSH public key: /home/ubuntu/.ssh/id_rsa.pub
Previewing changes that will be made:
I0216 17:14:51.332590 3547 apply_cluster.go:465] Gossip DNS: skipping DNS validation
I0216 17:14:51.392712 3547 executor.go:111] Tasks: 0 done / 83 total; 42 can run
W0216 17:14:51.792113 3547 vfs_castore.go:604] CA private key was not found
I0216 17:14:51.938057 3547 executor.go:111] Tasks: 42 done / 83 total; 17 can run
I0216 17:14:52.436407 3547 executor.go:111] Tasks: 59 done / 83 total; 18 can run
I0216 17:14:52.822395 3547 executor.go:111] Tasks: 77 done / 83 total; 2 can run
I0216 17:14:52.823088 3547 executor.go:111] Tasks: 79 done / 83 total; 2 can run
I0216 17:14:53.406919 3547 executor.go:111] Tasks: 81 done / 83 total; 2 can run
I0216 17:14:53.842148 3547 executor.go:111] Tasks: 83 done / 83 total; 0 can run
LaunchTemplate/master-eu-central-1a.masters.kOpsdemo1.k8s.local
AssociatePublicIP true
HTTPPutResponseHopLimit 1
HTTPTokens optional
IAMInstanceProfile name:masters.kOpsdemo1.k8s.local id:masters.kOpsdemo1.k8s.local
ImageID 099720109477/ubuntu/images/hvm-ssd/ubuntu-focal-20.04-amd64-server-20210119.1
InstanceType t2.micro
RootVolumeSize 64
RootVolumeType gp2
RootVolumeEncryption false
RootVolumeKmsKey
SSHKey name:kubernetes.kOpsdemo1.k8s.local-3e:19:92:ca:dd:64:d5:cf:ff:ed:3a:92:0f:40:d4:e8 id:kubernetes.kOpsdemo1.k8s.local-3e:19:92:ca:dd:64:d5:cf:ff:ed:3a:92:0f:40:d4:e8
SecurityGroups [name:masters.kOpsdemo1.k8s.local]
SpotPrice
Tags {k8s.io/cluster-autoscaler/node-template/label/kubernetes.io/role: master, k8s.io/cluster-autoscaler/node-template/label/kOps.k8s.io/instancegroup: master-eu-central-1a, k8s.io/role/master: 1, kOps.k8s.io/instancegroup: master-eu-central-1a, Name: master-eu-central-1a.masters.kOpsdemo1.k8s.local, KubernetesCluster: kOpsdemo1.k8s.local, kubernetes.io/cluster/kOpsdemo1.k8s.local: owned, k8s.io/cluster-autoscaler/node-template/label/node-role.kubernetes.io/master: }
Subnet/eu-central-1a.kOpsdemo1.k8s.local
ShortName eu-central-1a
VPC name:kOpsdemo1.k8s.local
AvailabilityZone eu-central-1a
CIDR 172.20.32.0/19
Shared false
Tags {KubernetesCluster: kOpsdemo1.k8s.local, kubernetes.io/cluster/kOpsdemo1.k8s.local: owned, SubnetType: Public, kubernetes.io/role/elb: 1, Name: eu-central-1a.kOpsdemo1.k8s.local}
VPC/kOpsdemo1.k8s.local
CIDR 172.20.0.0/16
EnableDNSHostnames true
EnableDNSSupport true
Shared false
Tags {kubernetes.io/cluster/kOpsdemo1.k8s.local: owned, Name: kOpsdemo1.k8s.local, KubernetesCluster: kOpsdemo1.k8s.local}
VPCDHCPOptionsAssociation/kOpsdemo1.k8s.local
VPC name:kOpsdemo1.k8s.local
DHCPOptions name:kOpsdemo1.k8s.local
Must specify --yes to apply changes
Cluster configuration has been created.
Suggestions:
* list clusters with: kOps get cluster
* edit this cluster with: kOps edit cluster kOpsdemo1.k8s.local
* edit your node instance group: kOps edit ig --name=kOpsdemo1.k8s.local nodes-eu-central-1a
* edit your master instance group: kOps edit ig --name=kOpsdemo1.k8s.local master-eu-central-1a
Finally configure your cluster with: kOps update cluster --name kOpsdemo1.k8s.local --yes –admin
Verify that the cluster was created.
ubuntu@ip-161-32-38-140:~$ kOps get cluster
NAME CLOUD ZONES
kOpsdemo1.k8s.local aws eu-central-1a
Apply the specified cluster specifications to the cluster.
ubuntu@ip-161-32-38-140:~$ kOps update cluster --name kOpsdemo1.k8s.local --yes --admin
I0216 17:15:24.800767 3556 apply_cluster.go:465] Gossip DNS: skipping DNS validation
I0216 17:15:25.919282 3556 executor.go:111] Tasks: 0 done / 83 total; 42 can run
W0216 17:15:26.343336 3556 vfs_castore.go:604] CA private key was not found
I0216 17:15:26.421652 3556 keypair.go:195] Issuing new certificate: "etcd-clients-ca"
I0216 17:15:26.450699 3556 keypair.go:195] Issuing new certificate: "etcd-peers-ca-main"
I0216 17:15:26.470785 3556 keypair.go:195] Issuing new certificate: "etcd-manager-ca-main"
I0216 17:15:26.531852 3556 keypair.go:195] Issuing new certificate: "etcd-peers-ca-events"
I0216 17:15:26.551601 3556 keypair.go:195] Issuing new certificate: "apiserver-aggregator-ca"
I0216 17:15:26.571834 3556 keypair.go:195] Issuing new certificate: "etcd-manager-ca-events"
I0216 17:15:26.592090 3556 keypair.go:195] Issuing new certificate: "master"
W0216 17:15:26.652894 3556 vfs_castore.go:604] CA private key was not found
I0216 17:15:26.653013 3556 keypair.go:195] Issuing new certificate: "ca"
I0216 17:15:31.344075 3556 executor.go:111] Tasks: 42 done / 83 total; 17 can run
I0216 17:15:32.306125 3556 executor.go:111] Tasks: 59 done / 83 total; 18 can run
I0216 17:15:34.189798 3556 executor.go:111] Tasks: 77 done / 83 total; 2 can run
I0216 17:15:34.190464 3556 executor.go:111] Tasks: 79 done / 83 total; 2 can run
I0216 17:15:35.738600 3556 executor.go:111] Tasks: 81 done / 83 total; 2 can run
I0216 17:15:37.810100 3556 executor.go:111] Tasks: 83 done / 83 total; 0 can run
I0216 17:15:37.904257 3556 update_cluster.go:313] Exporting kubecfg for cluster
kOps has set your kubectl context to kOpsdemo1.k8s.local
Cluster is starting. It should be ready in a few minutes.
Suggestions:
* validate cluster: kOps validate cluster --wait 10m
* list nodes: kubectl get nodes --show-labels
* ssh to the master: ssh -i ~/.ssh/id_rsa [email protected]
* the ubuntu user is specific to Ubuntu. If not using Ubuntu please use the appropriate user based on your OS.
* read about installing addons at: https://kOps.sigs.k8s.io/operations/addons.
If you immediately check the Kubernetes nodes are running or not, you will get an error. You need to be a little patient and wait for a few minutes (5-10) till the cluster is created.
ubuntu@ip-161-32-38-140:~$ kubectl get nodes
Unable to connect to the server: dial tcp: lookup api-kOpsdemo1-k8s-local-dason2-1001342368.eu-central-1.elb.amazonaws.com on 127.0.0.53:53: no such host
Validate the Cluster.
ubuntu@ip-161-32-38-140:~$ kOps validate cluster --wait 10m
Validating cluster kOpsdemo1.k8s.local
INSTANCE GROUPS
NAME ROLE MACHINETYPE MIN MAX SUBNETS
master-eu-central-1a Master t2.micro 1 1 eu-central-1a
nodes-eu-central-1a Node t2.micro 1 1 eu-central-1a
List the Nodes and Pods to check if all the nodes are ready and running. You can see both master and node are ready status.
ubuntu@ip-161-32-38-140:~$ kubectl get nodes
NAME STATUS ROLES AGE VERSION
ip-172-20-39-172.eu-central-1.compute.internal Ready master 11m v1.19.7
ip-172-20-56-248.eu-central-1.compute.internal Ready node 6m58s v1.19.7
You can check all the pods running in the Kubernetes cluster.
Just like creating a Kubernetes cluster, deleting a Kubernetes cluster also using kOps is very simple. This will remove all the cloud resources of the cluster and the cluster registry itself.
ubuntu@ip-161-32-38-140:~$ kOps delete cluster --name kOpsdemo1.k8s.local --yes
TYPE NAME ID
autoscaling-config master-eu-central-1a.masters.kOpsdemo1.k8s.local lt-0cc11aec1943204e4
autoscaling-config nodes-eu-central-1a.kOpsdemo1.k8s.local lt-0da65d2eaf6de9f5c
autoscaling-group master-eu-central-1a.masters.kOpsdemo1.k8s.local master-eu-central-1a.masters.kOpsdemo1.k8s.local
autoscaling-group nodes-eu-central-1a.kOpsdemo1.k8s.local nodes-eu-central-1a.kOpsdemo1.k8s.local
dhcp-options kOpsdemo1.k8s.local dopt-0403a0cbbfbc0c72b
iam-instance-profile masters.kOpsdemo1.k8s.local masters.kOpsdemo1.k8s.local
iam-instance-profile nodes.kOpsdemo1.k8s.local nodes.kOpsdemo1.k8s.local
iam-role masters.kOpsdemo1.k8s.local masters.kOpsdemo1.k8s.local
iam-role nodes.kOpsdemo1.k8s.local nodes.kOpsdemo1.k8s.local
instance master-eu-central-1a.masters.kOpsdemo1.k8s.local i-069c73f2c23eb502a
instance nodes-eu-central-1a.kOpsdemo1.k8s.local i-0401d6b0d4fc11e77
iam-instance-profile:nodes.kOpsdemo1.k8s.local ok
load-balancer:api-kOpsdemo1-k8s-local-dason2 ok
iam-instance-profile:masters.kOpsdemo1.k8s.local ok
iam-role:masters.kOpsdemo1.k8s.local ok
instance:i-069c73f2c23eb502a ok
autoscaling-group:nodes-eu-central-1a.kOpsdemo1.k8s.local ok
iam-role:nodes.kOpsdemo1.k8s.local ok
instance:i-0401d6b0d4fc11e77 ok
autoscaling-config:lt-0cc11aec1943204e4 ok
autoscaling-config:lt-0da65d2eaf6de9f5c ok
autoscaling-group:master-eu-central-1a.masters.kOpsdemo1.k8s.local ok
keypair:key-0e93f855f382a68fb ok
Deleted kubectl config for kOpsdemo1.k8s.local
Deleted cluster: "kOpsdemo1.k8s.local"
Conclusion
In this article, you’ve learned that kOps is a great tool to use for managing Kubernetes clusters hosted on various public cloud platforms because of the ways it can reduce the complexity. You’ve also learned the top kOps commands to know when working with Kubernetes, as well as and how to set up kubernetes containers on AWS.
Like this article?
Subscribe to our LinkedIn Newsletter to receive more educational content
We’re glad you are here! Densify customizes your experience by enabling cookies that help us understand your interests and recommend related information. By using our sites, you consent to our use of cookies. Learn more.
April 2, 2021 
