windows port

[maxsivkov]

This PR represent fluentd windows port

in order to run operator in heterogenous cluster (linux and windows nodes) one have to run two operator instances. One for linux and another for windows nodes.

Instructions

The easiest way to build linux and windows images is to have windows desktop (win 10) PC Of course you'll need Docker Desktop installed https://docs.docker.com/desktop/windows/install/. Docker Desktop should be running in WSL2 mode https://docs.docker.com/desktop/windows/wsl/ to be able to build linxu and windows images you need to install

• [Chocolatey] (https://chocolatey.org/install)

Set-ExecutionPolicy Bypass -Scope Process -Force; [System.Net.ServicePointManager]::SecurityProtocol = [System.Net.ServicePointManager]::SecurityProtocol -bor 3072; iex ((New-Object System.Net.WebClient).DownloadString('https://community.chocolatey.org/install.ps1'))

• Open cmd.exe in elevated prompt, install git (we will use bash from git package), make, jq

choco install git make jq -y

• Open cmd.exe, cd to the project root. Execute

"%ProgramFiles%\Git\git-bash.exe" --cd=.

• Open Docker Desktop, check if it is running in "Linux containers" mode. Build linux base image and reloader images with following command:

make build-image

• Push linux image with make push-image

• Switch Docker Desktop to "Windows containers" mode. Build windows images (for each windows platform you want to support https://hub.docker.com/_/microsoft-windows-servercore) NOTE: you need to change WIN_TAG for each iteration. For example: WIN_TAG := {ltsc2019, 1909, 2004, 20H2, ltsc2022}

  • Build base image for windows container with following command

    make build-windows-image IMG_OS=windows WIN_TAG=ltsc2019
    

    NOTE: you need to build as many base images as many windows versions you want to support (changing WIN_TAG parameter. Possible values https://hub.docker.com/_/microsoft-windows-servercore)

  • Build operator image

    make build-windows-image IMG_ARCH=amd64 IMG_OS=windows WIN_TAG=ltsc2019
    

    NOTE: you need to build as many base images as many windows versions you want to support (changing WIN_TAG parameter. Possible values https://hub.docker.com/_/microsoft-windows-servercore)

  • Push each windows image with make push-windows-image IMG_ARCH=amd64 IMG_OS=windows WIN_TAG=ltsc2019

• After all use make pushed-images to view all pushed images for the current release version

• Prepare multi-arch manifest

  • Execute make docker-manifest to create docker manifest and include all images for current release
  • Annotate each image with arch and os tags. I don't know how to automate this. For example

    IMAGE=vmware/kube-fluentd-operator
    VERSION=$(git describe --tags --always --dirty)
    
    # Annotate linux version
    docker manifest annotate $IMAGE:$VERSION $IMAGE:amd64-linux-$VERSION --arch amd64 --os linux
    # Annotate windows version
    docker manifest annotate $IMAGE:$VERSION $IMAGE:amd64-windows-ltsc2019-$VERSION --arch amd64 --os windows --os-version 10.0.17763.2114
    

  • Push manifest to the registry

    docker manifest push -p $IMAGE:$VERSION
    

• The same needs to be done in order to support 'latest' tag:

make docker-latest-manifest
IMAGE=vmware/kube-fluentd-operator
VERSION=$(git describe --tags --always --dirty)
# Annotate linux version
docker manifest annotate $IMAGE:latest $IMAGE:amd64-linux-$VERSION --arch amd64 --os linux
# Annotate windows version
docker manifest annotate $IMAGE:latest $IMAGE:amd64-windows-ltsc2019-$VERSION --arch amd64 --os windows --os-version 10.0.17763.2114
docker manifest push -p $IMAGE:latest

You may check if all images are combined under single manifest on dockerhub.com site. As an example: https://hub.docker.com/layers/168658076/maxsivkov/kube-fluentd-operator/latest/images/sha256-3da050dc7b041e9415a472e3e6c06c5835b65186aca8ac45d6b6952e2f12ad6a?context=repo

[vmwclabot]

@maxsivkov, you must sign every commit in this pull request acknowledging our Developer Certificate of Origin before your changes are merged. This can be done by adding Signed-off-by: John Doe <[email protected]> to the last line of each Git commit message. The e-mail address used to sign must match the e-mail address of the Git author. Click here to view the Developer Certificate of Origin agreement.

• Commit 46e58906e136ac990a088106fb5cc2597512f40a must be signed by [email protected]

[vmwclabot]

@maxsivkov, you must sign every commit in this pull request acknowledging our Developer Certificate of Origin before your changes are merged. This can be done by adding Signed-off-by: John Doe <[email protected]> to the last line of each Git commit message. The e-mail address used to sign must match the e-mail address of the Git author. Click here to view the Developer Certificate of Origin agreement.

• Commit 7bcd3d46f183e1117342d2dba50c3fce37e542ff must be signed by [email protected]

[Cryptophobia]

Hi @maxsivkov , thank you for the contributions! Looks good but I need to do a deeper review and test.

I will merge this in at some point after we make a new release and fix some of the config-reloader issues. I don't want to merge and create a regression in charts that we may need to chase later until we stabilize after the bug fixes.

[maxsivkov]

Hi @maxsivkov , thank you for the contributions! Looks good but I need to do a deeper review and test.

I will merge this in at some point after we make a new release and fix some of the config-reloader issues. I don't want to merge and create a regression in charts that we may need to chase later until we stabilize after the bug fixes.

Thanks. Please drop me a message when you'll be ready to process with my PR. I will update conflicting files.