kubenx is command line tool combining eks cli and kubectl
- Please Install
gofirst
# macos
$ brew tap GwonsooLee/gslee
$ brew install kubenx
$ kubenx version
Current Version is v1.0.0
# Linux
$ curl -LO https://feeldayone-public.s3.ap-northeast-2.amazonaws.com/release/linux/latest/kubenx.tar.gz
$ gzip -d kubenx.tar.gz
$ tar -xzvf kubenx.tar
$ mv kubenx /usr/local/bin
$ kubenx version
Current Version is v1.0.0- You can setup auto completion for kubenx command
# For Bash shell
echo "source <(kubenx completion bash)" >> ~/.bashrc
# For zsh shell
echo "source <(kubenx completion zsh)" >> ~/.zshrc- If you use multiple eks cluster in multiple AWS Account, then you need to assume role to use kubenx.
- Before using, it is
requiredto set up a configuration in$HOME/.kubenx/config.session_name: Role name you want to use for assume.assume: key-value mapping<Account Alias> : < Assume Role ARN >eks-assume-mapping: key-value mapping<Context> : <Account Alias>
- If you use
kubenx context, then it will automatically copy assume credentials to clipboard according to the configuration.- You should paste to shell by
Ctrl + v
- You should paste to shell by
{
"session_name": "Role name you want assume from",
"assume": {
"dev" : "arn:aws:iam::22222:role/role-name",
"stg" : "arn:aws:iam::33333:role/role-name",
"prod" : "arn:aws:iam::11111:role/role-name",
"security" : "arn:aws:iam::44444:role/role-name"
},
"eks-assume-mapping": {
"eks-prod-apnortheast2": "prod",
"eks-dev-apnortheast2": "dev",
"eks-stg-apnortheast2": "stg",
"eks-security-apnortheast2": "prod",
}
}$ kubenx context
Current Context: minikube
? Choose Context: [Use arrows to move, type to filter]
> minikube
docker-desktop
eks-benx-workshop
? Choose Context: docker-desktop
Context is changed to "docker-desktop"$ kubenx namespace
[ docker-desktop ] Current Namespace: argocd
? Choose Context: [Use arrows to move, type to filter]
> default
kube-node-lease
kube-public
kube-system
monitoring
Namespace is changed to "default"
## You can specify the namespace after command
$ kubenx ns kube-system
Namespace is changed to "kube-system"- If you want to get detail information about node, you can use this command
- By default, node filters applied is labels with
app,env. If you have these two labels on node, you could easily find the node when running inspect command. - You could get taint and pod information in the node you choose
- It will only search resources in
all namespaces. If you want to search in the specific namespace, please use-n <namespace>option.
$ kubenx inspect node
? Choose a node: [Use arrows to move, type to filter]
> docker-desktop (app=nginx, env=test)
? Choose a node: docker-desktop
========Taint INFO=======
purpose=common:NoSchedule
========POD INFO=======
NAME READY STATUS HOSTNAME POD IP HOST IP NODE AGE
nginx-deployment-56f8998dbc-5jvhr 1/1 Running 10.1.0.171 192.168.65.3 docker-desktop 31m
nginx-deployment-56f8998dbc-p8xnw 1/1 Running 10.1.0.172 192.168.65.3 docker-desktop 31m
nginx-deployment-56f8998dbc-pz4b2 1/1 Running 10.1.0.170 192.168.65.3 docker-desktop 32m
- You can search node and pod resource by label
- You should input
keyandvaluethrough shell and kubenx will search all nodes and pods with that label
? Key: app
? Value: nginx
Search Selector : app=nginx
========Node INFO=======
No node exists in the namespace
========POD INFO=======
NAME READY STATUS HOSTNAME POD IP HOST IP NODE AGE
nginx-deployment-56f8998dbc-5jvhr 1/1 Running 10.1.0.171 192.168.65.3 docker-desktop 33m
nginx-deployment-56f8998dbc-p8xnw 1/1 Running 10.1.0.172 192.168.65.3 docker-desktop 33m
nginx-deployment-56f8998dbc-pz4b2 1/1 Running 10.1.0.170 192.168.65.3 docker-desktop 34m
web-0 0/0 Pending web-0 34m
- You can clean configurations in kubeconfig.
- You can select multiple
contextby clickingspace key. - Of course you can search context while checking target cluster to delete.
$ kubenx config delete
? Pick contexts you want to delete: [Use arrows to move, space to select, type to filter]
[ ] eks-sample-apne2
[ ] eks-sample-apnortheast2-v2
> [x] minikube
[ ] eks-test-apnortheast2
[ ] eks-test2-apnortheast2
[ ] eks-common-k8s-useast2- You can update kubeconfig without searching eks cluster
$ kubenx config update
? Choose a cluster: [Use arrows to move, type to filter]
> eks-sample-apnortheast2-v1
eks-sample-apnortheast2-v2
Create new context eks-sample-apnortheast2-v1
## Also you can run with cluster name
$ kubenx config update eks-sample-apnortheast2-v1- Before you search clusters, please assume credentials of target AWS Account
- If you create new cluster with terraform, you need to add tag to VPC and subnets.
- Also you manually have to setup OIDC Provider in IAM.
- You can use
cluster initcommand which will do these work automatically
$ kubenx cluster init
******* Steps for initialization ********
Step 1. Tag setup for VPC
Step 2. Tag setup for public subnet
Step 3. Tag setup for private subnet
Step 4. Create Open ID Connector
? Choose a cluster: eks-stg-apnortheast2-v1
Step 1. VPC Tag needs to be updated
Step 2. Tags for Public Subnet needs to be updated
Step 3. Tags for Private Subnet is already updated
Step 4. New OIDC Provider is successfully created$ kubenx get cluster
+----------------------------------------------+-------------------------------------------------------------------------------+
| NAME | nginx_example_cluster |
+----------------------------------------------+-------------------------------------------------------------------------------+
| Version | 1.15 |
| Status | ACTIVE |
| Arn | arn:aws:eks:<Region ID>:<Account ID>:cluster/<Cluster Name> |
| Endpoint | <EKS ENDPOINT> |
| Cluster SG | <Security Group of Master> |
| VPC ID | <VPC Name>(<VPC ID>) |
| VPC Cidr Block | <VPC CIDR Block> |
| <Private Subent1 Name >(<subnet region>) | <subnet tag1 related to kubernetes> |
| | <subnet tag2 related to kubernetes> |
| | <subnet tag3 related to kubernetes> |
| | |
| <Private Subent2 Name >(<subnet region>) | <subnet tag1 related to kubernetes> |
| | <subnet tag2 related to kubernetes> |
| | <subnet tag3 related to kubernetes> |
+----------------------------------------------+-------------------------------------------------------------------------------+- You don't need to pass cluster name as argument. You can choose from the terminal
- If you want to pass cluster name as argument, you can specify it with [ --cluster, -c < cluster name > ]
$ kubenx get nodegroup
### Choose Nodegroup
? Choose a nodegroup: [Use arrows to move, type to filter]
> eks-nginx-node-group1
eks-nginx-node-group2
eks-nginx-node-group3
### You can see the information about Nodegroup
? Choose a nodegroup: eks-nginx-node-group1
NAME STATUS INSTANCE TYPE LABELS MIN SIZE DISIRED SIZE MAX SIZE AUTOSCALING GROUPDS DISK SIZE
eks-nginx-node-group1 ACTIVE t3.small app=nginx,env=dev 1 1 1 eks-d8b88e2f-75c2-03c4-6c99-b54b6ad02312 20
AUTOSCALING GROUP INSTANCE ID HEALTH STATUS INSTANCE TYPE AVAILABILITY ZONE
eks-d8b88e2f-75c2-03c4-6c99-b54b6ad02312 i-041412f0s19f24b1b Healthy t3.small ap-northeast-2cKubectl Command
$ kubectl get pod
NAME READY STATUS RESTARTS AGE
nginx-deployment-56f8998dbc-5jvhr 1/1 Running 0 8m15s
nginx-deployment-56f8998dbc-p8xnw 1/1 Running 0 8m14s
nginx-deployment-56f8998dbc-pz4b2 1/1 Running 0 8m46s
web-0 0/1 Pending 0 8m40sKubenx Command
- You can see
Pod IP,Host IP, andthe node it is scheduled.
$ kubenx get pod
NAME READY STATUS HOSTNAME POD IP HOST IP NODE AGE
nginx-deployment-56f8998dbc-5jvhr 1/1 Running 10.1.0.171 192.168.65.3 docker-desktop 8m33s
nginx-deployment-56f8998dbc-p8xnw 1/1 Running 10.1.0.172 192.168.65.3 docker-desktop 8m32s
nginx-deployment-56f8998dbc-pz4b2 1/1 Running 10.1.0.170 192.168.65.3 docker-desktop 9m4s
web-0 0/0 Pending web-0 8m58sKubectl Command
$ kubectl get service
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
kubernetes ClusterIP 10.96.0.1 <none> 443/TCP 4d8h
nginx ClusterIP None <none> 80/TCP 8m31sKubenx Command
- You can find
endpoint listwhich the service is routing to.
$ kubenx get service
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) ENDPOINT(S) AGE
kubernetes ClusterIP 10.96.0.1 <None> 443/TCP 10.1.0.171,10.1.0.172,10.1.0.170, 4d8h
nginx ClusterIP None <None> 80/TCP 10.1.0.171,10.1.0.172,10.1.0.170, 8m7sKubectl Command
$ kubectl get deployment
NAME READY UP-TO-DATE AVAILABLE AGE
nginx-deployment 3/3 3 3 14mKubenx Command
- You can find the
deployment strategyand the configurations about it.
$ kubenx get deployment
NAME READY UP-TO-DATE AVAILABLE STRATEGY TYPE MAXUNAVAILABLE MAXSURGE CONTAINERS IMAGE AGE
nginx-deployment 3 3 3 RollingUpdate 25% 25% nginx nginx:1.9.1 14mKubectl Command
$ kubectl get ingresses.
NAME HOSTS ADDRESS PORTS AGE
ingress-nginx * 80 7m53sKubenx Command
- All ports and paths are shown.
- You can see
target servicein the list.
$ kubenx get ingress
NAME HOST ADDRESS PATH PORTS TARGET SERVICE AGE
ingress-nginx * /* 80 service-nginx 7m48sKubectl Command
$ kubectl get pods
NAME READY STATUS RESTARTS AGE
nginx-5578584966-czg6j 1/1 Running 0 100m
$ kubectl port-forward nginx-5578584966-czg6j 8099:80
Forwarding from 127.0.0.1:8099 -> 80
Forwarding from [::1]:8099 -> 80
Kubenx Command
- You are able to choose pod from terminal.
- You can cancel anytime via
<Ctrl> + c. - Default pod port is the local port you choose.
$ kubenx port-forward
? Choose a pod: [Use arrows to move, type to filter]
> nginx-5578584966-czg6j
? Choose a pod: nginx-5578584966-czg6j
? Local port to use: 8080
? Pod port[ Default: 8080]: 80
Forwarding from 127.0.0.1:8080 -> 80
Forwarding from [::1]:8080 -> 80
Port forwarding is ready to get traffic. have fun!