How can I make EKS kick working with CircleCI? - docker

Question

I can push images to ECR but I am not even close sure what I should do next (what should be the flow) to make my images run on Kubernetes on EKS
jobs:
create-deployment:
executor: aws-eks/python3
parameters:
cluster-name:
description: |
Name of the EKS cluster
type: string
steps:
- checkout
- aws-eks/update-kubeconfig-with-authenticator:
cluster-name: << parameters.cluster-name >>
install-kubectl: true
- kubernetes/create-or-update-resource:
get-rollout-status: true
resource-file-path: tests/nginx-deployment/deployment.yaml
# resource-file-path: configs/k8s/prod-deployment.yaml
resource-name: deployment/prod-deployment
orbs:
aws-ecr: circleci/aws-ecr#6.15.0
aws-eks: circleci/aws-eks#1.1.0
kubernetes: circleci/kubernetes#0.4.0
version: 2.1
workflows:
deployment:
jobs:
- aws-ecr/build-and-push-image:
repo: bwtc-backend
tag: "${CIRCLE_BRANCH}-v0.1.${CIRCLE_BUILD_NUM}"
dockerfile: configs/Docker/Dockerfile.prod
path: .
filters:
branches:
ignore:
- master
- aws-eks/create-cluster:
cluster-name: eks-demo-deployment
requires:
- aws-ecr/build-and-push-image
- create-deployment:
cluster-name: eks-demo-deployment
requires:
- aws-eks/create-cluster
- aws-eks/update-container-image:
cluster-name: eks-demo-deployment
container-image-updates: 'nginx=nginx:1.9.1'
post-steps:
- kubernetes/delete-resource:
resource-names: nginx-deployment
resource-types: deployment
wait: true
record: true
requires:
- create-deployment
resource-name: deployment/nginx-deployment
- aws-eks/delete-cluster:
cluster-name: eks-demo-deployment
requires:
- aws-eks/update-container-image
That's what I've got in my config for now.
The problem I am facing at the moment is:
gzip: stdin: not in gzip format
tar: Child returned status 1
tar: Error is not recoverable: exiting now
Exited with code exit status 2
CircleCI received exit code 2
I am using a snippet from CircleCI Documentation, so I guess it should work.
I passed in all the params as I can see but I can't get what I've missed here.
I need your help guys!

Answer 1

The last update has a bug. It is failing due to an incorrect URL. The issue is still open, see here. The correct url should be
https://github.com/weaveworks/eksctl/releases/latest/download/eksctl_$(uname -s)_amd64.tar.gz
unlike the current
"https://github.com/weaveworks/eksctl/releases/download/latest_release/eksctl_$(uname -s)_amd64.tar.gz"
While you wait for it to be approved, use the following to install eksctl first.
- run:
name: Install the eksctl tool
command: |
if which eksctl > /dev/null; then
echo "eksctl is already installed"
exit 0
fi
mkdir -p eksctl_download
curl --silent --location --retry 5 "https://github.com/weaveworks/eksctl/releases/latest/download/eksctl_$(uname -s)_amd64.tar.gz" \
| tar xz -C eksctl_download
chmod +x eksctl_download/eksctl
SUDO=""
if [ $(id -u) -ne 0 ] && which sudo > /dev/null ; then
SUDO="sudo"
fi
$SUDO mv eksctl_download/eksctl /usr/local/bin/
rmdir eksctl_download
and then run the job
- aws-eks/create-cluster:
cluster-name: eks-demo-deployment
That should solve the issue.
An example
# Creation of Cluster
create-cluster:
executor: aws-eks/python3
parameters:
cluster-name:
description: |
Name of the EKS cluster
type: string
steps:
- run:
name: Install the eksctl tool
command: |
if which eksctl > /dev/null; then
echo "eksctl is already installed"
exit 0
fi
mkdir -p eksctl_download
curl --silent --location --retry 5 "https://github.com/weaveworks/eksctl/releases/latest/download/eksctl_$(uname -s)_amd64.tar.gz" \
| tar xz -C eksctl_download
chmod +x eksctl_download/eksctl
SUDO=""
if [ $(id -u) -ne 0 ] && which sudo > /dev/null ; then
SUDO="sudo"
fi
$SUDO mv eksctl_download/eksctl /usr/local/bin/
rmdir eksctl_download
- aws-eks/create-cluster:
cluster-name: eks-demo-deployment