-
Story
-
Resolution: Done
-
Major
-
None
-
3
-
False
-
-
False
-
RHOAISTRAT-30 - MVP support for Kubeflow Pipelines v2 in data science pipelines
-
No
-
No
-
-
-
RHOAI DOCS 2.9, RHOAI DOCS 2.9 CONTINUED, RHOAI DOCS 2.10
-
Testable
Version 2 of Data Science Pipelines is being released at RHOAI 2.9. As part of this release, the backing DSP engine is being changed from Tekton to Argo Workflows. As part of the documentation changes for DSP v2, we need to remove all Tekton content from the documentation.
As part of these changes, we must rewrite this module regarding running distributed workflows using data science pipelines.
If you have a look at the module using the link, you will notice the mention of the kfp-tekton dependency in step 3i. Also, in the code example in step 3iii, we have code that relates to the Tekton compiler, which will also need to be replaced with the appropriate code for the KubeFlow Pipelines instead.
This module needs to be rewritten to fully comply with the changes required for Data Science Pipelines Version 2, due to be introduced in RHOAI 2.9.
For further information on what these changes should consist of, see the following slack thread:
https://redhat-internal.slack.com/archives/C05U1Q749PV/p1709220835054549
- is related to
-
RHOAIENG-1621 integration test cases with DSP, CodeFlare and KubeRay in a multi-step pipeline
- Closed
- relates to
-
RHOAIENG-2999 [ccs] Remove all Tekton-based documentation
- Closed
- mentioned on