Support for the aarch64 arm64 architecture

yashprakashsharma commented 3 weeks ago

Validation Checklist

[X] Is this a Kubeflow issue?
[X] Are you posting in the right repository ?
[X] Did you follow the installation guide https://github.com/kubeflow/manifests?tab=readme-ov-file ?
[X] Is the issue report properly structured and detailed with version numbers?
[ ] Is this for Kubeflow development ?
[ ] Would you like to work on this issue?
[ ] Join our slack channel using wg-manifests.

Version

1.8

Describe your issue

kubeflow ran on my old x86_64 architecture linux instance but using same method for a new instance with aarch64(arm64) arch multiple pods ceash cause containers in them fail continuously on researching further i found images for failing container were not supporting arm64 arch. Images for reference can be found here: https://kubeflow.slack.com/archives/C7REE0ETX/p1717502674581769

Steps to reproduce the issue

kubeflow manifest v1.8.0 installation on aarch64

Put here any screenshots or videos (optional)

https://kubeflow.slack.com/archives/C7REE0ETX/p1717502674581769

juliusvonkohout commented 1 week ago

Please try with Kubeflow 1.9 https://github.com/kubeflow/manifests/commits/v1.9-branch or the master branch

juliusvonkohout commented 1 week ago

Merged https://github.com/kubeflow/manifests/issues/2659 into this issue

juliusvonkohout commented 4 days ago

@yashprakashsharma you need to retry with the 1.9 release and post your findings here, not in slack.

yashprakashsharma commented 4 days ago

Hi @juliusvonkohout, I already migrated to x86_64 architecture, and I am using v1.8.1 it works fine, I no longer have access to a arm64 machine to try v1.9. I will give it a try whenever I get access to arm64 machine

juliusvonkohout commented 4 days ago

Well if anyone finds any images not available for arm64 in the Kubeflow 1.9 branch please post them here.

AndersBennedsgaard commented 3 days ago

KFP does still not run on arm64: https://github.com/kubeflow/pipelines/issues/10309 Katib also doesn't: https://github.com/kubeflow/katib/issues/2049 It might not be relevant for Kubeflow, but KServe also doesn't: https://github.com/kserve/models-web-app/issues/73

kubeflow / manifests