Terraform/Kubernetes Reference Infrastructure for Cloud Posse Production Organization in AWS.
NOTE: Before creating the Production infrastructure, you need to provision the Parent ("Root") Organization in AWS (because it creates resources needed for all other accounts). Follow the steps in README first. You need to do it only once.
This project is part of our comprehensive "SweetOps" approach towards DevOps.
It's 100% Open Source and licensed under the APACHE2.
We use geodesic to define and build world-class cloud infrastructures backed by AWS and powered by Kubernetes.
geodesic exposes many tools that can be used to define and provision AWS and Kubernetes resources.
Here is the list of tools we use to provision cloudposse.co infrastructure:
NOTE: You need to do it only once.
Configure AWS profile in ~/.aws/config. Make sure to change username (username@cloudposse.co) to your own.
[profile cpco-prod-admin]
region=us-west-2
role_arn=arn:aws:iam::845778104613:role/OrganizationAccountAccessRole
mfa_serial=arn:aws:iam::323330167063:mfa/username@cloudposse.co
source_profile=cpcoNOTE: You need to do it only once.
We use aws-vault to store IAM credentials in your operating system's secure keystore and then generates temporary credentials from those to expose to your shell and applications.
Install aws-vault on your local computer first.
On MacOS, you may use homebrew cask
brew cask install aws-vaultThen setup your secret credentials for AWS in aws-vault
aws-vault add --backend file cpcoNOTE: You should set AWS_VAULT_BACKEND=file in your shell rc config (e.g. ~/.bashrc) so it persists.
For more info, see aws-vault
# Initialize the project's build-harness
make init
# Build docker image
make docker/build
make installprod.cloudposse.coassume-roleNOTE: You need to do it only once for a given set of secrets. Repeat this step if you want to add new secrets.
Populate chamber secrets for kops project (make sure to change the keys and values to reflect your environment; add new secrets as needed)
chamber write kops <key1> <value1>
chamber write kops <key2> <value2>
...NOTE: Run chamber list -e kops to list the secrets stored for kops project
Populate chamber secrets for backing-services project (make sure to change the values to reflect your environment; add new secrets as needed)
chamber write backing-services TF_VAR_POSTGRES_ADMIN_NAME admin
chamber write backing-services TF_VAR_POSTGRES_ADMIN_PASSWORD supersecret
chamber write backing-services TF_VAR_POSTGRES_DB_NAME appNOTE: Run chamber list -e backing-services to list the secrets stored for backing-services project
NOTE: Before provisioning AWS resources with Terraform, you need to create tfstate-backend first (S3 bucket to store Terraform state and DynamoDB table for state locking).
Follow the steps in this README. You need to do it only once.
After tfstate-backend has been provisioned, follow the rest of the instructions in the order shown below.
Change directory to dns folder
cd /conf/dnsRun Terraform
init-terraform
terraform plan
terraform applyFor more info, see geodesic-with-terraform
cd /conf/cloudtrail
init-terraform
terraform plan
terraform applycd /conf/acm
init-terraform
terraform plan
terraform applycd /conf/chamber
init-terraform
terraform plan
terraform applyAvailable targets:
all Initialize build-harness, install deps, build docker container, install wrapper script and run shell
build Build docker image
deps Install dependencies (if any)
help This help screen
help/all Display help for all targets
install Install wrapper script from geodesic container
push Push docker image to registry
run Start the geodesic shell by calling wrapper script
We create a kops cluster from a manifest.
The manifest template is located in /templates/kops/default.yaml
and is compiled by running build-kops-manifest in the Dockerfile.
Provisioning a kops cluster takes three steps:
- Provision the
kopsbackend (config S3 bucket, cluster DNS zone, and SSH keypair to access the k8s masters and nodes) in Terraform - Update the
Dockerfileand rebuild/restart thegeodesicshell to generate akopsmanifest file - Execute the
kopsmanifest file to create thekopscluster
Change directory to kops folder
cd /conf/kopsRun Terraform to provision the kops backend (S3 bucket, DNS zone, and SSH keypair)
init-terraform
terraform plan
terraform applyFrom the Terraform outputs, copy the zone_name and bucket_name into the ENV vars KOPS_CLUSTER_NAME and KOPS_STATE_STORE in the Dockerfile.
The Dockerfile kops config should look like this:
# kops config
ENV KOPS_CLUSTER_NAME="us-west-2.prod.cloudposse.co"
ENV KOPS_DNS_ZONE=${KOPS_CLUSTER_NAME}
ENV KOPS_STATE_STORE="s3://cpco-prod-kops-state"
ENV KOPS_STATE_STORE_REGION="us-west-2"
ENV KOPS_AVAILABILITY_ZONES="us-west-2a,us-west-2b,us-west-2c"
ENV KOPS_BASTION_PUBLIC_NAME="bastion"
ENV BASTION_MACHINE_TYPE="t2.medium"
ENV MASTER_MACHINE_TYPE="t2.large"
ENV NODE_MACHINE_TYPE="t2.large"
ENV NODE_MAX_SIZE="3"
ENV NODE_MIN_SIZE="3"
Type exit (or hit ^D) to leave the shell.
Note, if you've assumed a role, you'll first need to leave that also by typing exit (or hit ^D).
Rebuild the Docker image
NOTE: We provision backing-services in two phases because:
aurora-postgresandelasticache-redisdepend onvpc-peering(they usekopsSecurity Group to allowkopsapplications to connect)vpc-peeringdepends onvpcandkops(it creates a peering connection between the two networks)
To break the circular dependencies, we provision kops, then vpc (from backing-services), then vpc-peering,
and finally the rest of backing-services (aurora-postgres and elasticache-redis).
NOTE: We use chamber to first populate the environment with the secrets from the specified service (backing-services)
and then execute the given commands (terraform plan and terraform apply)
cd /conf/backing-services
init-terraform
chamber exec backing-services -- terraform plan -target=module.identity -target=module.vpc -target=module.subnets
chamber exec backing-services -- terraform apply -target=module.identity -target=module.vpc -target=module.subnetscd /conf/kops-aws-platform
init-terraform
terraform plan -target=module.identity -target=data.aws_vpc.backing_services_vpc -target=module.kops_vpc_peering
terraform apply -target=module.identity -target=data.aws_vpc.backing_services_vpc -target=module.kops_vpc_peeringcd /conf/kops-aws-platform
terraform plan
terraform applyNOTE: We use chamber to first populate the environment with the secrets from the specified service (backing-services)
and then execute the given commands (terraform plan and terraform apply)
cd /conf/backing-services
chamber exec backing-services -- terraform plan
chamber exec backing-services -- terraform applyWe use helmfile to deploy Helm charts to provision Kubernetes resources.
helmfile.yaml is located in the /conf/kops directory in geodesic container (see helmfile.yaml).
Change the current directory to kops
cd /conf/kopsDeploy the Helm charts
NOTE: We use chamber to first populate the environment with the secrets from the kops service and then execute the given command (helmfile sync)
kops export kubecfg $KOPS_CLUSTER_NAME
chamber exec kops -- helmfile sync✅ (cpco-prod-admin) kops ➤ chamber exec kops -- helmfile sync
exec: helm repo add stable https://kubernetes-charts.storage.googleapis.com
"stable" has been added to your repositories
exec: helm repo add cloudposse-incubator https://charts.cloudposse.com/incubator/
"cloudposse-incubator" has been added to your repositories
exec: helm repo update
Hang tight while we grab the latest from your chart repositories...
...Skip local chart repository
...Successfully got an update from the "incubator" chart repository
...Successfully got an update from the "cloudposse-incubator" chart repository
...Successfully got an update from the "stable" chart repository
...Successfully got an update from the "coreos-stable" chart repository
Update Complete. ⎈ Happy Helming!⎈
exec: helm upgrade --install kube2iam stable/kube2iam --version 0.8.5 --namespace kube-system --set tolerations[0].key=node-role.kubernetes.io/master,tolerations[0].effect=NoSchedule,aws.region=us-west-2,extraArgs.auto-discover-base-arn=true,host.iptables=true,host.interface=cali+,resources.limits.cpu=4m,resources.limits.memory=16Mi,resources.requests.cpu=4m,resources.requests.memory=16Mi
exec: helm upgrade --install kube-lego cloudposse-incubator/kube-lego --version 0.1.2 --namespace kube-system --values /localhost/Documents/Projects/CloudPosse/Programs/Projects/Joany/prod.cloudposse.co/conf/kops/values/kube-lego.yaml
exec: helm upgrade --install ingress cloudposse-incubator/nginx-ingress --version 0.1.7 --namespace kube-system --values /localhost/Documents/Projects/CloudPosse/Programs/Projects/Joany/prod.cloudposse.co/conf/kops/values/ingress.yaml
exec: helm upgrade --install external-dns stable/external-dns --version 0.5.4 --namespace kube-system --set nodeSelector.kubernetes\.io/role=master,extraEnv.EXTERNAL_DNS_SOURCE=service
ingress,tolerations[0].key=node-role.kubernetes.io/master,tolerations[0].effect=NoSchedule,txtOwnerId=us-west-2.prod.cloudposse.co,txtPrefix=184f3df5-53c6-4071-974b-2d8de32e82c7-,publishInternalServices=true,provider=aws,podAnnotations.iam\.amazonaws\.com/role=cpco-prod-external-dns,resources.limits.cpu=100m,resources.limits.memory=128Mi,resources.requests.cpu=100m,resources.requests.memory=128Mi
exec: helm upgrade --install chart-repo cloudposse-incubator/chart-repo --version 0.2.1 --namespace kube-system --values /localhost/Documents/Projects/CloudPosse/Programs/Projects/Joany/prod.cloudposse.co/conf/kops/values/chart-repo.yaml
Release "kube-lego" does not exist. Installing it now.
NAME: kube-lego
LAST DEPLOYED: Wed Apr 18 14:46:47 2018
NAMESPACE: kube-system
STATUS: DEPLOYED
RESOURCES:
==> v1/ConfigMap
NAME DATA AGE
kube-lego-kube-lego 2 1s
==> v1beta1/Deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
kube-lego-kube-lego 1 1 1 0 1s
==> v1/Pod(related)
NAME READY STATUS RESTARTS AGE
kube-lego-kube-lego-d88d9c968-kr94m 0/1 ContainerCreating 0 1s
NOTES:
Kube Lego has been installed to the kube-system as kube-lego-kube-lego.
Lego endpoint: https://acme-v01.api.letsencrypt.org/directory
Release "kube2iam" does not exist. Installing it now.
NAME: kube2iam
LAST DEPLOYED: Wed Apr 18 14:46:47 2018
NAMESPACE: kube-system
STATUS: DEPLOYED
RESOURCES:
==> v1beta1/DaemonSet
NAME DESIRED CURRENT READY UP-TO-DATE AVAILABLE NODE SELECTOR AGE
kube2iam 5 5 0 5 0 <none> 2s
==> v1/Pod(related)
NAME READY STATUS RESTARTS AGE
kube2iam-754dm 0/1 ContainerCreating 0 2s
kube2iam-95mz8 0/1 ContainerCreating 0 2s
kube2iam-klhtc 0/1 ContainerCreating 0 2s
kube2iam-m9v5z 0/1 ContainerCreating 0 2s
kube2iam-xvkvt 0/1 ContainerCreating 0 2s
NOTES:
To verify that kube2iam has started, run:
kubectl --namespace=kube-system get pods -l "app=kube2iam,release=kube2iam"
Add an iam.amazonaws.com/role annotation to your pods with the role you want them to assume.
https://github.com/jtblin/kube2iam#kubernetes-annotation
Use `curl` to verify the pod's role from within:
curl http://169.254.169.254/latest/meta-data/iam/security-credentials/
Release "chart-repo" does not exist. Installing it now.
NAME: chart-repo
E0418 14:46:49.898191 478 portforward.go:303] error copying from remote stream to local connection: readfrom tcp4 127.0.0.1:46183->127.0.0.1:41002: write tcp4 127.0.0.1:46183->127.0.0.1:41002: write: broken pipe
LAST DEPLOYED: Wed Apr 18 14:46:47 2018
NAMESPACE: kube-system
STATUS: DEPLOYED
RESOURCES:
==> v1/Secret
NAME TYPE DATA AGE
chart-repo-gateway Opaque 2 1s
chart-repo-server Opaque 2 1s
==> v1/Service
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
chart-repo-gateway ClusterIP 100.68.33.126 <none> 8080/TCP 1s
chart-repo-server ClusterIP 100.64.249.2 <none> 8080/TCP 1s
==> v1beta1/Deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
chart-repo-gateway 2 2 2 0 1s
chart-repo-server 2 2 2 0 1s
==> v1beta1/Ingress
NAME HOSTS ADDRESS PORTS AGE
chart-repo-gateway gateway.charts.us-west-2.prod.cloudposse.co 80, 443 1s
chart-repo-server charts.us-west-2.prod.cloudposse.co 80, 443 1s
==> v1/Pod(related)
NAME READY STATUS RESTARTS AGE
chart-repo-gateway-b947dd69-jgf7c 0/1 ContainerCreating 0 1s
chart-repo-gateway-b947dd69-v8n6n 0/1 ContainerCreating 0 1s
chart-repo-server-d447dfdb6-4bbzn 0/1 ContainerCreating 0 1s
chart-repo-server-d447dfdb6-kfsl8 0/1 ContainerCreating 0 1s
NOTES:
Thank you for installing chart-repo.
Your release is named chart-repo.
Release "ingress" does not exist. Installing it now.
NAME: ingress
LAST DEPLOYED: Wed Apr 18 14:46:48 2018
NAMESPACE: kube-system
STATUS: DEPLOYED
RESOURCES:
==> v1/ConfigMap
NAME DATA AGE
ingress-nginx-default-ba-config 1 2s
ingress-nginx-default-ba 2 2s
ingress-nginx-ingress 2 2s
==> v1/Service
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
ingress-nginx-default-ba ClusterIP 100.67.227.42 <none> 80/TCP 2s
ingress-nginx-ingress LoadBalancer 100.69.186.114 <pending> 80:32182/TCP,443:30549/TCP 2s
==> v1beta1/Deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
ingress-nginx-default-ba 2 2 2 0 2s
ingress-nginx-ingress 4 4 4 0 2s
==> v1/Pod(related)
NAME READY STATUS RESTARTS AGE
ingress-nginx-default-ba-f6fd8b978-fj2zx 0/1 ContainerCreating 0 2s
ingress-nginx-default-ba-f6fd8b978-jx9t6 0/1 ContainerCreating 0 2s
ingress-nginx-ingress-76bc4ff7cc-4knmh 0/1 ContainerCreating 0 2s
ingress-nginx-ingress-76bc4ff7cc-4mplg 0/1 ContainerCreating 0 2s
ingress-nginx-ingress-76bc4ff7cc-8rwqz 0/1 ContainerCreating 0 2s
ingress-nginx-ingress-76bc4ff7cc-96v82 0/1 ContainerCreating 0 2s
Release "external-dns" does not exist. Installing it now.
NAME: external-dns
LAST DEPLOYED: Wed Apr 18 14:46:48 2018
NAMESPACE: kube-system
STATUS: DEPLOYED
RESOURCES:
==> v1/Secret
NAME TYPE DATA AGE
external-dns Opaque 3 1s
==> v1/Service
NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
external-dns ClusterIP 100.68.153.195 <none> 7979/TCP 1s
==> v1beta1/Deployment
NAME DESIRED CURRENT UP-TO-DATE AVAILABLE AGE
external-dns 1 1 1 0 1s
==> v1/Pod(related)
NAME READY STATUS RESTARTS AGE
external-dns-7bb969cf47-xdxzf 0/1 ContainerCreating 0 1s
NOTES:
To verify that external-dns has started, run:
kubectl --namespace=kube-system get pods -l "app=external-dns,release=external-dns"
Verify that all deployed Kubernetes resources are up and running
✅ (cpco-prod-admin) kops ➤ kube-system get pods
NAME READY STATUS RESTARTS AGE
calico-kube-controllers-6b5f557d7d-lm2x7 1/1 Running 0 9d
calico-node-5txbt 2/2 Running 0 9d
calico-node-9mpgm 2/2 Running 0 9d
calico-node-cx777 2/2 Running 0 9d
calico-node-gcswp 2/2 Running 0 9d
calico-node-nqmch 2/2 Running 0 9d
chart-repo-gateway-b947dd69-jgf7c 1/1 Running 2 1m
chart-repo-gateway-b947dd69-v8n6n 1/1 Running 2 1m
chart-repo-server-d447dfdb6-4bbzn 1/1 Running 2 1m
chart-repo-server-d447dfdb6-kfsl8 1/1 Running 2 1m
dns-controller-6ddf5d44d5-l92c8 1/1 Running 0 9d
etcd-server-events-ip-172-20-127-167.us-west-2.compute.internal 1/1 Running 0 9d
etcd-server-events-ip-172-20-38-15.us-west-2.compute.internal 1/1 Running 0 9d
etcd-server-events-ip-172-20-73-251.us-west-2.compute.internal 1/1 Running 0 9d
etcd-server-ip-172-20-127-167.us-west-2.compute.internal 1/1 Running 0 9d
etcd-server-ip-172-20-38-15.us-west-2.compute.internal 1/1 Running 0 9d
etcd-server-ip-172-20-73-251.us-west-2.compute.internal 1/1 Running 0 9d
external-dns-7bb969cf47-xdxzf 1/1 Running 0 1m
ingress-nginx-default-ba-f6fd8b978-fj2zx 1/1 Running 0 1m
ingress-nginx-default-ba-f6fd8b978-jx9t6 1/1 Running 0 1m
ingress-nginx-ingress-76bc4ff7cc-4knmh 1/1 Running 0 1m
ingress-nginx-ingress-76bc4ff7cc-4mplg 1/1 Running 0 1m
ingress-nginx-ingress-76bc4ff7cc-8rwqz 1/1 Running 0 1m
ingress-nginx-ingress-76bc4ff7cc-96v82 1/1 Running 0 1m
kube-apiserver-ip-172-20-127-167.us-west-2.compute.internal 1/1 Running 0 9d
kube-apiserver-ip-172-20-38-15.us-west-2.compute.internal 1/1 Running 0 9d
kube-apiserver-ip-172-20-73-251.us-west-2.compute.internal 1/1 Running 2 9d
kube-controller-manager-ip-172-20-127-167.us-west-2.compute.internal 1/1 Running 0 9d
kube-controller-manager-ip-172-20-38-15.us-west-2.compute.internal 1/1 Running 0 9d
kube-controller-manager-ip-172-20-73-251.us-west-2.compute.internal 1/1 Running 0 9d
kube-dns-7f56f9f8c7-62hbv 3/3 Running 0 9d
kube-dns-7f56f9f8c7-kq2c9 3/3 Running 0 9d
kube-dns-autoscaler-f4c47db64-hcss2 1/1 Running 0 9d
kube-lego-kube-lego-d88d9c968-kr94m 1/1 Running 0 1m
kube-proxy-ip-172-20-127-167.us-west-2.compute.internal 1/1 Running 0 9d
kube-proxy-ip-172-20-38-15.us-west-2.compute.internal 1/1 Running 0 9d
kube-proxy-ip-172-20-43-225.us-west-2.compute.internal 1/1 Running 0 9d
kube-proxy-ip-172-20-73-251.us-west-2.compute.internal 1/1 Running 0 9d
kube-proxy-ip-172-20-89-216.us-west-2.compute.internal 1/1 Running 0 9d
kube-scheduler-ip-172-20-127-167.us-west-2.compute.internal 1/1 Running 0 9d
kube-scheduler-ip-172-20-38-15.us-west-2.compute.internal 1/1 Running 0 9d
kube-scheduler-ip-172-20-73-251.us-west-2.compute.internal 1/1 Running 0 9d
kube2iam-754dm 1/1 Running 0 1m
kube2iam-95mz8 1/1 Running 0 1m
kube2iam-klhtc 1/1 Running 0 1m
kube2iam-m9v5z 1/1 Running 0 1m
kube2iam-xvkvt 1/1 Running 0 1m
tiller-deploy-f44659b6c-dn6p2 1/1 Running 0 9d
Check out these related projects.
- Packages - Cloud Posse installer and distribution of native apps
- Build Harness - Collection of Makefiles to facilitate building Golang projects, Dockerfiles, Helm charts, and more
- terraform-root-modules - Collection of Terraform "root module" invocations for provisioning reference architectures
- root.cloudposse.co - Example Terraform Reference Architecture of a Geodesic Module for a Parent ("Root") Organization in AWS.
- audit.cloudposse.co - Example Terraform Reference Architecture of a Geodesic Module for an Audit Logs Organization in AWS.
- staging.cloudposse.co - Example Terraform Reference Architecture of a Geodesic Module for a Staging Organization in AWS.
- dev.cloudposse.co - Example Terraform Reference Architecture of a Geodesic Module for a Development Sandbox Organization in AWS.
- testing.cloudposse.co - Example Terraform Reference Architecture of a Geodesic Module for a Testing Organization in AWS.
For additional context, refer to some of these links.
- Reference Architecture Documentation - Futher explanation of how to use these reference architectures
- Cloud Posse Documentation - Complete documentation for the Cloud Posse solution
- Chamber - Chamber is a CRUD tool for managing secrets stored in AWS Systems Manager Parameter Store and exposing those secrets as Environment Variables to processes.
- The Right Way to Store Secrets using Parameter Store - Centrally and securely manage secrets with Amazon EC2 Systems Manager Parameter Store, lots of Terraform code, and chamber. This post has all the information you need to get running with Parameter Store in production.
- external-dns - Frequently asked Questions & Answers
- Kubernetes Production Patterns - Explore helpful techniques to improve resiliency and high availability of Kubernetes deployments and will take a look at some common mistakes to avoid when working with Docker and Kubernetes.
Got a question?
File a GitHub issue, send us an email or join our Slack Community.
Work directly with our team of DevOps experts via email, slack, and video conferencing.
We provide commercial support for all of our Open Source projects. As a Dedicated Support customer, you have access to our team of subject matter experts at a fraction of the cost of a fulltime engineer.
- Questions. We'll use a Shared Slack channel between your team and ours.
- Troubleshooting. We'll help you triage why things aren't working.
- Code Reviews. We'll review your Pull Requests and provide constructive feedback.
- Bug Fixes. We'll rapidly work to fix any bugs in our projects.
- Build New Terraform Modules. We'll develop original modules to provision infrastructure.
- Cloud Architecture. We'll assist with your cloud strategy and design.
- Implementation. We'll provide hands on support to implement our reference architectures.
Get access to our Open Source Community Forum on Slack. It's FREE to join for everyone! Our "SweetOps" community is where you get to talk with others who share a similar vision for how to rollout and manage infrastructure. This is the best place to talk shop, ask questions, solicit feedback, and work together as a community to build sweet infrastructure.
Please use the issue tracker to report any bugs or file feature requests.
If you are interested in being a contributor and want to get involved in developing this project or help out with our other projects, we would love to hear from you! Shoot us an email.
In general, PRs are welcome. We follow the typical "fork-and-pull" Git workflow.
- Fork the repo on GitHub
- Clone the project to your own machine
- Commit changes to your own branch
- Push your work back up to your fork
- Submit a Pull Request so that we can review your changes
NOTE: Be sure to merge the latest changes from "upstream" before making a pull request!
Copyright © 2017-2018 Cloud Posse, LLC
See LICENSE for full details.
Licensed to the Apache Software Foundation (ASF) under one
or more contributor license agreements. See the NOTICE file
distributed with this work for additional information
regarding copyright ownership. The ASF licenses this file
to you under the Apache License, Version 2.0 (the
"License"); you may not use this file except in compliance
with the License. You may obtain a copy of the License at
https://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing,
software distributed under the License is distributed on an
"AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
KIND, either express or implied. See the License for the
specific language governing permissions and limitations
under the License.
All other trademarks referenced herein are the property of their respective owners.
This project is maintained and funded by Cloud Posse, LLC. Like it? Please let us know at hello@cloudposse.com
We're a DevOps Professional Services company based in Los Angeles, CA. We love Open Source Software!
We offer paid support on all of our projects.
Check out our other projects, apply for a job, or hire us to help with your cloud strategy and implementation.