Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

remove minioJob #2331

Merged
merged 1 commit into from
Oct 16, 2024
Merged

remove minioJob #2331

merged 1 commit into from
Oct 16, 2024

Conversation

jiuker
Copy link
Contributor

@jiuker jiuker commented Oct 8, 2024

remove minioJob

remove minioJob
@pjuarezd
Copy link
Member

pjuarezd commented Oct 8, 2024

remove minioJob

@ramondeklein @dvaldivia @jiuker @harshavardhana a bit of info on why this feature is being remove would be helpful

@ramondeklein
Copy link
Contributor

@pjuarezd I think it's considered an enterprise feature.

@pjuarezd pjuarezd merged commit 6cd2041 into minio:master Oct 16, 2024
21 checks passed
@nicolasps-wisemodo
Copy link

I just wanted to mention, I installed the operator from scratch using the main branch, using the default values.yaml file, with terrible results. The operator was no longer functional, tenants were not being created.

On the operator's logs, the following errors appeared several times present:

failed to watch *v1alpha1.miniojob: failed to list *v1alpha1.miniojob: the server could not find the requested resource (get miniojobs.job.min.io)

waiting for the sts certificates secret to be issued

Initially, I thought it was some kind of RBAC issue causing the error, but then I realized the MinioJob CRD was missing. I looked for the missing CRD in version v6.0.4 (https://github.com/minio/operator/blob/v6.0.4/helm/operator/templates/job.min.io_jobs.yaml) and copied it into operator/helm/operator/templates/, and after restarting the operator's pod, everything worked as expected. The error messages disappeared from the log, and for the first time the tenant was created.

@ramondeklein
Copy link
Contributor

@nicolasps-wisemodo Please use tagged releases.

@nicolasps-wisemodo
Copy link

No doubt I will @ramondeklein

@dharmil
Copy link

dharmil commented Nov 2, 2024

Does this mean the future versions won't support MinIOJob? That kind of contradicts the advice given in #1100

@Syndlex
Copy link

Syndlex commented Nov 4, 2024

@ramondeklein, @pjuarezd I understand that we don’t have entitlement to specific features.

However, I believe this is a crucial feature for any Minio deployment, regardless of the use case. While it's possible to implement a workaround, it's not particularly difficult, which may reduce the incentive to move to the Enterprise version.

For smaller companies or opensource projects, creating these workarounds can be time-consuming, and there are already several options available, such as #1100, InseeFrLab S3 Operator, creating a Kubernetes Job, using Helm hooks or ArgoCD hooks.

Could you kindly provide some clarity on why this feature is categorized as an Enterprise feature?

@ramondeklein
Copy link
Contributor

We have evaluated MinioJob and there are a few issues with it.

The community actually requests a fully declarative way (so MinioBucket, MinioUser, MinioPolicy, ... CRDs). We have evaluated this idea, but we need to create an almost endless amount of CRDs to match the S3 API. Re-using mc was way simpler. When we would have created a CRD-based approach, then others would probably ask for a Hashicorp Terraform provider. It's a rabit hole...

Each functionality requires development, but that's not all. We need to maintain, document and also support it. There are other projects that do supply such functionality (like https://github.com/InseeFrLab/s3-operator). We think we should focus on what is important. Adding all kind of niche-features distracts us from doing the most important part, building an high-performance, scalable and reliable object store.

For these reasons we have decided to remove this feature. There are other methods that can be used to initiate MinIO jobs (like spinning up an mc job yourself) or you can use alternative methods.

@Syndlex
Copy link

Syndlex commented Nov 7, 2024

@ramondeklein thank you for clarification. This makes total sense and I am happy with this answer.

I think AWS should donate S3 to the CNCF or something so this can become a proper Standard.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

6 participants