airshipctl/docs/source/developers.md
James Gu 06609c3180 Added script for the missing controlplane_target phase
controlplane_target phase is needed for target cluster more than a
single node

Signed-off-by: James Gu <james.gu@att.com>
Change-Id: I6e77d4268cdee0ebcc65e1f9172ef645ced53337
2021-04-13 10:47:47 -07:00

17 KiB

Developer's Guide

This guide explains how to set up your environment for developing on airshipctl.

Environment expectations

  • Git
  • Go 1.13
  • Docker

Installing Git

Instructions to install Git are here.

Installing Go 1.13

Instructions to install Golang are here.

The make test verification step requires the GNU Compiler Collection (gcc) to be installed.

To install the GNU Compiler Collection (gcc):

sudo apt-get install gcc

Proxy Setup

If your organization requires development behind a proxy server, you will need to define the following environment variables with your organization's information:

HTTP_PROXY=http://username:password@host:port
HTTPS_PROXY=http://username:password@host:port
NO_PROXY="localhost,127.0.0.1,10.23.0.0/16,10.96.0.0/12"
PROXY=http://username:password@host:port
USE_PROXY=true

10.23.0.0/16 encapsulates the range of addresses used by airshipctl development environment virtual machines, and 10.96.0.0/12 is the Kubernetes service CIDR.

DNS Configuration

If you cannot reach the Google DNS servers from your local environment, add your DNS servers to manifests/type/airship-core/shared/catalogues/common-networking.yaml in place of the Google ones.

Clone airshipctl code

Run the following command to download the latest airshipctl code:

git clone https://opendev.org/airship/airshipctl.git

NOTE: The airshipctl application is a Go module. This means that there is no need to clone the repository into the $GOPATH directory in order to build it. You should be able to build it from any directory as long as $GOPATH is defined correctly.

Installing Docker & Other Tools

Prior to building the airshipctl binary, ensure you have Docker, Ansible & other tools installed in your environment.

There is a script in the airshipctl directory named 00_setup.sh which can be run to download all the required binaries and packages. This script code can be viewed here.

Standalone instructions to install Docker are here. This is not necessary if you run 00_setup.sh.

Building airshipctl

Run the following command to build the airshipctl binary:

make build

This will compile airshipctl and place the resulting binary into the bin directory.

To test the build, including linting and coverage reports, run:

make test

To run all tests in a containerized environment, run:

make docker-image-test-suite

Docker Images

To build an airshipctl Docker image, run:

make docker-image

Pre-built images are already available at quay.io. Moreover, in the directory airshipctl/tools/gate/, different scripts are present which will run and download all the required images. The script 10_build_gate.sh will download all the required images.

Contribution Guidelines

We welcome contributions. This project has set up some guidelines in order to ensure that

  • code quality remains high
  • the project remains consistent, and
  • contributions follow the open source legal requirements.

Our intent is not to burden contributors, but to build elegant and high-quality open source code so that our users will benefit. Make sure you have read and understood the main airshipctl Contributing Guide.

Structure of the Code

The code for the airshipctl project is organized as follows:

  • The client-facing code is located in cmd/. Code inside of cmd/ is not designed for library reuse.
  • Shared libraries are stored in pkg/.
  • Both commands and shared libraries may require test data fixtures. These should be placed in a testdata/ subdirectory within the command or library.
  • The testutil/ directory contains functions that are helpful for unit tests.
  • The zuul.d/ directory contains Zuul YAML definitions for CI/CD jobs to run.
  • The playbooks/ directory contains playbooks that the Zuul CI/CD jobs will run.
  • The tools/ directory contains scripts used by the Makefile and CI/CD pipeline.
  • The tools/gate directory consists of different scripts. These scripts will setup the environment as per requirements and install all the required packages and binaries. This will also download all the required docker images.
  • The docs/ folder is used for documentation and examples.
  • Go dependencies are managed by go mod and stored in go.mod and go.sum

Git Conventions

We use Git for our version control system. The master branch is the home of the current development candidate. Releases are tagged. We accept changes to the code via Gerrit pull requests. One workflow for doing this is as follows:

  1. git clone the airshipctl repository. For this run the command:

    git clone https://opendev.org/airship/airshipctl.git
    
  2. Use OpenDev documentation to setup Gerrit with the repo.

  3. When set, use this guide to learn the OpenDev development workflow, in a sandbox environment. You can then apply the learnings to start developing airshipctl.

Go Conventions

We follow the Go coding style standards very closely. Typically, running goimports -w -local opendev.org/airship/airshipctl ./ will make your code beautiful for you.

We also typically follow the conventions of golangci-lint. Read more:

Testing

In order to ensure that all package unit tests follow the same standard and use the same frameworks, airshipctl has a document outlining specific test guidelines maintained separately. Moreover, there are few scripts in directory tools/gate which run different tests. The script 20_run_gate_runner.sh will generate airshipctl config file, deploy ephemeral cluster with infra and cluster API, deploy target cluster and verify all control pods.

Steps to build a Local All-in-one VM Environment

Pre-requisites: Make sure the following conditions are met:

  1. Nested Virtualization enabled on the Host

  2. A Virtual Machine with 20 GB RAM, 4 vCPU and 80GB Disk and Ubuntu 18.04 Installed.

  3. Clone the following repo -

  4. Install necessary packages and pre deployment setup

    1. ./tools/gate/00_setup.sh
  5. Download test security key and add it to environment variable.

    - curl -fsSL -o /tmp/key.asc https://raw.githubusercontent.com/mozilla/sops/master/pgp/sops_functional_tests_key.asc
    - export SOPS_IMPORT_PGP="$(cat /tmp/key.asc)"
    - export SOPS_PGP_FP="FBC7B9E2A4F9289AC0C1D4843D16CEE4A27381B4"
    
  6. Execute the following scripts one by one

    1. ./tools/gate/10_build_gate.sh
    2. ./tools/gate/20_run_gate_runner.sh

    Note: instead of running ./tools/gate/20_run_gate_runner.sh it's possible to run its steps one-by-one, from playbooks/airshipctl-gate-runner.yaml, e.g.:

    1. sudo -E ./tools/deployment/01_install_kubectl.sh
    2. sudo -E ./tools/deployment/22_test_configs.sh
    3. ...
  7. How to verify the ephemeral cluster and target cluster is deployed successfully Validate Ephemeral Cluster is Operational:

    kubectl --kubeconfig /home/user/.airship/kubeconfig --context ephemeral-cluster get pods --all-namespaces
    NAMESPACE                           NAME                                                             READY   STATUS    RESTARTS   AGE
    capi-kubeadm-bootstrap-system       capi-kubeadm-bootstrap-controller-manager-556678c94-hngzj        2/2     Running   0          50s
    capi-kubeadm-control-plane-system   capi-kubeadm-control-plane-controller-manager-556d47dffd-qljht   2/2     Running   0          47s
    capi-system                         capi-controller-manager-67859f6b78-2tgcx                         2/2     Running   0          54s
    capi-webhook-system                 capi-controller-manager-5c785c685c-fds47                         2/2     Running   0          55s
    capi-webhook-system                 capi-kubeadm-bootstrap-controller-manager-77658d7745-5bb7z       2/2     Running   0          52s
    capi-webhook-system                 capi-kubeadm-control-plane-controller-manager-74dcf8b9c-ds4l7    2/2     Running   0          49s
    capi-webhook-system                 capm3-controller-manager-568747bbbb-zld5v                        2/2     Running   0          45s
    capm3-system                        capm3-controller-manager-698c6d6df9-n72cf                        2/2     Running   0          42s
    cert-manager                        cert-manager-578cd6d964-lznfq                                    1/1     Running   0          76s
    cert-manager                        cert-manager-cainjector-5ffff9dd7c-h9v6l                         1/1     Running   0          76s
    cert-manager                        cert-manager-webhook-556b9d7dfd-hvvfs                            1/1     Running   0          75s
    hardware-classification             hardware-classification-controller-manager-776b5f66f8-6z9xl      2/2     Running   0          10m
    kube-system                         calico-kube-controllers-94b8f9766-6cl6l                          1/1     Running   0          10m
    kube-system                         calico-node-dw6c8                                                1/1     Running   0          10m
    kube-system                         coredns-66bff467f8-57wpm                                         1/1     Running   0          13m
    kube-system                         coredns-66bff467f8-lbfw2                                         1/1     Running   0          13m
    kube-system                         etcd-ephemeral                                                   1/1     Running   0          13m
    kube-system                         kube-apiserver-ephemeral                                         1/1     Running   0          13m
    kube-system                         kube-controller-manager-ephemeral                                1/1     Running   0          13m
    kube-system                         kube-proxy-whdhw                                                 1/1     Running   0          13m
    kube-system                         kube-scheduler-ephemeral                                         1/1     Running   0          13m
    metal3                              ironic-5d95b49d6c-lr6b2                                          4/4     Running   0          10m
    metal3                              metal3-baremetal-operator-84f9df77fb-zq4qv                       3/3     Running   0          10m
    

    Validate Target Cluster is Operational:

    kubectl --kubeconfig /home/user/.airship/kubeconfig --context target-cluster get pods --all-namespaces
    NAMESPACE                           NAME                                                             READY   STATUS    RESTARTS   AGE
    capi-kubeadm-bootstrap-system       capi-kubeadm-bootstrap-controller-manager-556678c94-svqmn        2/2     Running   0          56s
    capi-kubeadm-control-plane-system   capi-kubeadm-control-plane-controller-manager-556d47dffd-z28lq   2/2     Running   0          46s
    capi-system                         capi-controller-manager-67859f6b78-x4k25                         2/2     Running   0          64s
    capi-webhook-system                 capi-controller-manager-5c785c685c-9t58p                         2/2     Running   0          69s
    capi-webhook-system                 capi-kubeadm-bootstrap-controller-manager-77658d7745-wv8bt       2/2     Running   0          62s
    capi-webhook-system                 capi-kubeadm-control-plane-controller-manager-74dcf8b9c-rskqk    2/2     Running   0          51s
    capi-webhook-system                 capm3-controller-manager-568747bbbb-gpvqc                        2/2     Running   0          35s
    capm3-system                        capm3-controller-manager-698c6d6df9-n6pfm                        2/2     Running   0          27s
    cert-manager                        cert-manager-578cd6d964-nkgj7                                    1/1     Running   0          99s
    cert-manager                        cert-manager-cainjector-5ffff9dd7c-ps62z                         1/1     Running   0          99s
    cert-manager                        cert-manager-webhook-556b9d7dfd-2spgg                            1/1     Running   0          99s
    flux-system                         helm-controller-cbb96fc8d-7vh96                                  1/1     Running   0          11m
    flux-system                         source-controller-64f4b85496-zfj6w                               1/1     Running   0          11m
    hardware-classification             hardware-classification-controller-manager-776b5f66f8-zd5rt      2/2     Running   0          11m
    kube-system                         calico-kube-controllers-94b8f9766-9r2cn                          1/1     Running   0          11m
    kube-system                         calico-node-6gfpc                                                1/1     Running   0          11m
    kube-system                         coredns-66bff467f8-4gggz                                         1/1     Running   0          16m
    kube-system                         coredns-66bff467f8-qgbhj                                         1/1     Running   0          16m
    kube-system                         etcd-node01                                                      1/1     Running   0          16m
    kube-system                         kube-apiserver-node01                                            1/1     Running   0          16m
    kube-system                         kube-controller-manager-node01                                   1/1     Running   0          16m
    kube-system                         kube-proxy-ch6z9                                                 1/1     Running   0          16m
    kube-system                         kube-scheduler-node01                                            1/1     Running   0          16m
    metal3                              ironic-5d95b49d6c-8xwcx                                          4/4     Running   0          11m
    metal3                              metal3-baremetal-operator-84f9df77fb-25h4w                       3/3     Running   0          11m
    
  8. How to deploy Workloads Once the Target is Operational, Workloads can be deployed on the Target Cluster. A small demo workload can be deployed using ./tools/deployment/36_deploy_workload.sh.This demo includes ingress as a workload. To verify execute kubectl command as below:

    $ kubectl --kubeconfig /home/user/.airship/kubeconfig --context target-cluster get pods -n ingress
    
    NAME                                                    READY   STATUS    RESTARTS   AGE
    ingress-ingress-nginx-controller-7d5d89f47d-p8hms       1/1     Running   1          6d19h
    ingress-ingress-nginx-defaultbackend-6c49f4ff7f-nzsjw   1/1     Running   1          6d19h
    

    Additional Workloads can be defined under ~/airshipctl/manifests/site/test-site/target/workload/kustomization.yaml which specifies the resources as below

    $ pwd
    /home/user/airshipctl/manifests/site/test-site/target/workload
    $ cat kustomization.yaml
    resources:
    - ../../../../function/airshipctl-base-catalogues
    - ../../../../type/gating/target/workload
    transformers:
    - ../../../../type/gating/target/workload/ingress/replacements
    $ pwd
    /home/user/airshipctl/manifests/type/gating/target/workload
    $ ll
    total 16
    drwxrwxr-x 3 user user 4096 Nov 16 17:02 ./
    drwxrwxr-x 3 user user 4096 Nov 16 17:02 ../
    drwxrwxr-x 3 user user 4096 Nov 16 17:02 ingress/
    -rw-rw-r-- 1 user user   23 Nov 16 17:02 kustomization.yaml
    
  9. In case the All-in-One-VM is restarted and the nested VMs do not get restarted automatically simply execute the below steps to make the Target Cluster up again.

    $ sudo virsh list --all
    Id    Name                           State
    ----------------------------------------------------
    -     air-ephemeral                  shut off
    -     air-target-1                   shut off
    -     air-worker-1                   shut off
    $ virsh net-start air_nat
    Network air_nat started
    $ virsh net-start air_prov
    Network air_prov started
    $ virsh start air-target-1
    Domain air-target-1 started
    $ virsh start air-worker-1
    Domain air-worker-1 started
    $ sudo virsh list --all
    Id    Name                           State
    ----------------------------------------------------
    3     air-target-1                   running
    4     air-worker-1                   running
    
  10. In case the deployment needs to be cleaned and rerun again, run the below script.

    • sudo ./tools/deployment/clean.sh