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.

Sign up for GitHub

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

Jump to bottom

Users should be able to add metadata to component.yaml #1013

Closed
animeshsingh opened this issue Mar 21, 2019 · 2 comments · Fixed by #1077
Closed

Users should be able to add metadata to component.yaml #1013

animeshsingh opened this issue Mar 21, 2019 · 2 comments · Fixed by #1077
Assignees
@Ark-kun
Labels
area/sdk/components

Comments

@animeshsingh
Copy link
Contributor

currently it appears components are checked for exact match on the schema, but this can be made extensible by giving an option to pass addition attributes etc which a user may define, which is not exactly used by pipeline subsystem, but can be information which an entity can use to display things e.g. link to logo etc...
@Ark-kun Ark-kun self-assigned this Mar 22, 2019
@Ark-kun
Copy link
Contributor

Ark-kun commented Mar 22, 2019

I think that having a strict schema is important and the general structure needs to be fixed, so th the components are portable and we can maintain backwards compatibility.

But.

I think that the cases you've mentioned can be covered by adding the metadata section to the component.yaml

In keeping it similar to kubernetes and Argo, we can have annotations and labels inside.

Would that fit your needs?

@Ark-kun Ark-kun changed the title components yaml should be extensible Users should be able to add metadata to component.yaml Mar 22, 2019
@Ark-kun Ark-kun mentioned this issue Mar 22, 2019
Merged
@Ark-kun
Copy link
Contributor

Ark-kun commented Mar 22, 2019

Check the #1023 PR and /lgtm if it fulfills your need.

Ark-kun added a commit to Ark-kun/pipelines that referenced this issue Apr 3, 2019
@Ark-kun
Passing the annotations and labels to the ContainerOp
beb31f5 
Currently the annotations and labels are not passed from component to the ContainerOp. This PR fixes that.

Fixes kubeflow#1013
@Ark-kun Ark-kun mentioned this issue Apr 3, 2019
Merged
Ark-kun added a commit to Ark-kun/pipelines that referenced this issue Apr 9, 2019
@Ark-kun
Passing the annotations and labels to the ContainerOp
addec1e 
Currently the annotations and labels are not passed from component to the ContainerOp. This PR fixes that.

Fixes kubeflow#1013
k8s-ci-robot pushed a commit that referenced this issue Apr 9, 2019
@Ark-kun @k8s-ci-robot
Passing the annotations and labels to the ContainerOp (#1077)
929ff52 
Currently the annotations and labels are not passed from component to the ContainerOp. This PR fixes that.

Fixes #1013
Linchin pushed a commit to Linchin/pipelines that referenced this issue Apr 11, 2023
@chensun
kfp: Reduce persistent agent TTL_SECONDS_AFTER_WORKFLOW_FINISH to 8 h…
6e7467e 
…ours (kubeflow#1013)
magdalenakuhn17 pushed a commit to magdalenakuhn17/pipelines that referenced this issue Oct 22, 2023
@yuzliu
TrainedModel controller (kubeflow#1013)
52451bd 
* Trained model api (kubeflow#1009)

* trained model api

* remove frameworks' version information

trainedmodel related code

add groupversion_info

generate deepcopy

add v1beta1 serving api

remove InferenceService depedency

fix optional comment

rename MultiModelConfigMapName to ModelConfigMapName and rename
tainedmodel to tm

rename multiModelConfigMapName to modelConfigName

initialize trainedmodel controller

rename multimodelconfig pkg to modelconfig

refactor shardingstrategy

refactor strategy

generate crd

* rebase master

* update modelconfig when reconciling TrainedModel

* create an instance of modelconfig reconciler in trainedmodel controller's loop otherwise the reconciler is a nil pointer

* add unit test

* rename

* rename

* fix unit tests

* revert changes for v1alpha2/openapi_generated.go

* remove config/crd/serving.kubeflow.org_inferencerouters.yaml

* rename MultiModelConfigVolumeName to ModelConfigVolumeName

* rename trainedmodel's predictorModel to inference

* inline error check

* fix log

* annotate error

* remove unnecessary error check

* rename package

* remove unit test for delete as it does not trigger a delete event

* refactor sharding strategy

* rename configmap to modelConfig

* rename ConfigMapReconciler to ModelConfigReconciler

* fmt

* pass modelconfig reconciler to TrainedModelReconciler as a field

* rename TrainedModelSpec.Inference to TrainedModelSpec.model

* change Result{Requeue: true, RequeueAfter: 10 * time.Second} to Result{} since the framework will requeue when err is not nil

* fix plural name

* fix description

* fix unit test name

* add argo retry

* changne retry to 3

* add comma

* revert makefile so it only deploys v1alpha1

* register v1beta1 scheme

* fix trainedmodel rbac

* fix trainedmodel crd per kubernetes-retired/cluster-registry#255

* fix trainedmodel crd from makefile

* run make manifests

* clean up

* fix comment

* install kfserving sdk to user home folder
HumairAK pushed a commit to red-hat-data-services/data-science-pipelines that referenced this issue Mar 11, 2024
@Tomcli
Update result name to be case sensitive (kubeflow#1013)
eaac53d 
* update result name to be case sensitive

* add custom task test case
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees

@Ark-kun Ark-kun

Labels
area/sdk/components
Projects
None yet
Milestone
No milestone
Development

Successfully merging a pull request may close this issue.

Passing the annotations and labels to the ContainerOp Ark-kun/pipelines
2 participants
@Ark-kun @animeshsingh