ArgoCD deployment

[Syntax3rror404]

How can I install kubeflow with ArgoCD?

[jbottum]

/kind feature request /area installation /priority p2

@Syntax3rror404 thanks for your request. Unless I am missing something, this pattern is not supported i.e. as you probably have seen, the community has developed a kustomize installation pattern. there is some work on helm but that is not developed or supported. We can bring this request up in the Community meeting but it probably is a question for the Manifest Working group meetings. I suspect an Argo CD installation pattern would be a XXL sized feature, but willing to see what the community says. @kubeflow/wg-manifests-leads

[google-oss-prow[bot]]

@jbottum: The label(s) kind/request, area/installation cannot be applied, because the repository doesn't have them.

In response to [this](https://github.com/kubeflow/manifests/issues/2374#issuecomment-1412248542): >/kind feature request >/area installation >/priority p2 > >@Syntax3rror404 thanks for your request. Unless I am missing something, this pattern is not supported i.e. as you probably have seen, the community has developed a kustomize installation pattern. there is some work on helm but that is not developed or supported. We can bring this request up in the Community meeting but it probably is a question for the Manifest Working group meetings. I suspect an Argo CD installation pattern would be a XXL sized feature, but willing to see what the community says. @kubeflow/wg-manifests-leads Instructions for interacting with me using PR comments are available [here](https://git.k8s.io/community/contributors/guide/pull-requests.md). If you have questions or suggestions related to my behavior, please file an issue against the [kubernetes/test-infra](https://github.com/kubernetes/test-infra/issues/new?title=Prow%20issue:) repository.

[Syntax3rror404]

A helm installation option would be perfect and would make the installation much easier. Helm is state of the art for installing kubernetes apps. Helm apps can also deployed with argo.

[jbottum]

You might check here: https://github.com/kubeflow/kubeflow/issues/3173

[thesuperzapper]

@Syntax3rror404 you will be very interested in the new version of Argoflow that @Talador12 and I are working on.

The current repo is quite messy and only supports Kubeflow 1.3, but we have a working version for 1.6+ that we are testing.

[Syntax3rror404]

@thesuperzapper I also seen argoflow, but there is no commit since two years. Why is it so imposible to install kubeflow right now? For normal applications I can set up a values file and fire helm against my cluster and the app comes up?

[agilgur5]

So ArgoCD does support Kustomize. Several teams I work with have used that pattern to install it. It is pretty complicated to set-up nonetheless, especially if you have some of the dependencies already in your cluster and configured (e.g. we use Argo Workflows and Istio independently of Kubeflow).

Better organization around this and a Helm chart would certainly be preferable! I'd like to contribute a Helm chart(s) per my comment in https://github.com/kubeflow/kubeflow/issues/3173#issuecomment-1622742413 🙂 Would think that would belong here in the manifests repo

[thesuperzapper]

So...

Thanks for everyone's patience. Over the last few years I have been building deployKF, it's effectively a reimagined version of Argoflow, but it's also SO much more than that too.

I am very excited for you all to try it, a crazy amount of work has been required, spanning multiple years, but I think it will completely change how open-source ML platforms are built on Kubernetes.

The first public beta will very likely drop tomorrow (but please don't get too mad if it takes a few more days, you will understand when you get to try it).

❤️ ❤️ ❤️

[thesuperzapper]

Hi everyone, I am very excited to share the first release of deployKF!

Try it out here: https://github.com/deployKF/deployKF

Why you should care.

With deployKF, it's WAY easier to configure things, because everything has a config value (no more manual patches):

• defining static user logins
• connecting with identity providers like Microsoft AD, Okta, GitHub, Google Workspace, etc
• defining users/profiles/groups
• connecting Kubeflow Pipelines with S3
• connecting Kubeflow Pipelines with an existing MySQL db
• configuring Kubeflow Notebooks with custom images
• exposing the main istio gateway as a LoadBalancer or with an Ingress
• etc.

We have patched LOTS of security issues, many of which can't be easily fixed in the upstream manifests:

• All secrets are randomly generated at install time, rather than being hardcoded in manifests.
• Reduced attack vectors compared to Kubeflow Manifests, particularly in Istio configurations.
• Uses Istio with distroless images by default.
• MinIO (or S3) access keys are isolated to each profile, and scoped to the minimum required permissions.
• Supports using AWS IRSA instead of S3 access keys.
• (some redacted stuff)

The whole design of deployKF is focused on creating enterprise-ready ML Platforms which are actually possible to maintain in the long term, we have:

• In-place upgrades (bring your old custom values files to new versions, we aim to not break backward compatibility)
• Changes to configs/secrets will automatically restart the affected services
• GitOps with the industry standard, ArgoCD (note, support for other CD tools will come, and are even making a native installer that does not require GitOps)

deployKF is more than just Kubeflow, we aim to support all leading open ML & Data tools, for example:

• MLflow Model Registry
• Apache Airfow
• Feast
• (so many more, see the full list here)

❤️ deployKF has been a long labor of love, with work spanning back at least 3 years! I am so excited to see how everyone uses it, and how the industry reacts to the best ML Platform being a free and open-source tool!

[Syntax3rror404]

deployKF are not solving my problem because I need a helm chart which is default and best practise in kubernetes.

[thesuperzapper]

Thanks for reminding me about this issue.

As I believe deployKF resolves the issue around deploying Kubeflow with ArgoCD, I will be closing this issue for now. If anyone in the future creates an alternative solution, please feel free to raise it here.

Also, I would also like to draw everyone's attention to the Kubeflow and CNCF codes of conduct, as consistently breaking them may result in being banned from this repo.

/close

[google-oss-prow[bot]]

@thesuperzapper: Closing this issue.

In response to [this](https://github.com/kubeflow/manifests/issues/2374#issuecomment-1826196176): >Thanks for reminding me about this issue. > >As I believe [deployKF](https://www.deploykf.org/) resolves the issue around deploying Kubeflow with ArgoCD, I will be closing this issue for now. If anyone in the future creates an alternative solution, please feel free to raise it here. > >Also, I would also like to draw everyone's attention to the [Kubeflow](https://github.com/kubeflow/community/blob/master/CODE_OF_CONDUCT.md) and [CNCF](https://github.com/cncf/foundation/blob/main/code-of-conduct.md) codes of conduct, as consistently breaking them may result in being banned from this repo. > >/close Instructions for interacting with me using PR comments are available [here](https://git.k8s.io/community/contributors/guide/pull-requests.md). If you have questions or suggestions related to my behavior, please file an issue against the [kubernetes/test-infra](https://github.com/kubernetes/test-infra/issues/new?title=Prow%20issue:) repository.