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

[RayCluster][Feature] add GcsFaultToleranceOptions to the RayCluster CRD [2/N] #2720

Closed
2 tasks done
rueian opened this issue Jan 8, 2025 · 2 comments · Fixed by #2721, #2731 or #2760
Closed
2 tasks done

[RayCluster][Feature] add GcsFaultToleranceOptions to the RayCluster CRD [2/N] #2720

rueian opened this issue Jan 8, 2025 · 2 comments · Fixed by #2721, #2731 or #2760
Assignees
Labels
1.3.0 enhancement New feature or request

Comments

@rueian
Copy link
Contributor

rueian commented Jan 8, 2025

Search before asking

  • I had searched in the issues and found no similar feature requirement.

Description

Currently, enabling RayCluster GCS FT is cumbersome given that users need to update many settings that are scattered around in their CR yaml files. The new GcsFaultToleranceOptions field in RayCluster CRD has been merged (#2715). The field serves as a new central place containing all configs for enabling GCS FT. Now we need to implement the actual logic to enable the feature according to the new field.

Our goal is when the new field is set, the KubeRay should configure the RayCluster with GCS FT enabled. That includes setting corresponding Pods annotations, environment variables, and RayStartParams. See the doc for more details.

We plan to implement it in 2 separate PRs:

  1. Set RayFTEnabledAnnotationKey and RayExternalStorageNSAnnotationKey pod annotations according to the GcsFaultToleranceOptions.
  2. Set the RAY_REDIS_ADDRESS environment variable to Head Pod according to the GcsFaultToleranceOptions.
  3. Set the REDIS_PASSWORD environment variable and the corresponding RayStartParam to Head Pod according to the GcsFaultToleranceOptions.
  4. Set the REDIS_USERNAME environment variable and the corresponding RayStartParam to Head Pod according to the GcsFaultToleranceOptions as well as adding the support of the REDIS_USERNAME to the usage of the cleanup_redis_storage.

This issue targets the first part: Setting pod annotations.

Related issues

#2695

Are you willing to submit a PR?

  • Yes I am willing to submit a PR!
@rueian
Copy link
Contributor Author

rueian commented Jan 9, 2025

#2721 has addressed the 1. and 2. parts.

@fscnick
Copy link
Contributor

fscnick commented Jan 12, 2025

#2731 has addressed the 3. parts.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment