Merge pull request kubeflow#3205 from lluunn/v05-kfctl051

kfctl use 0.5.1

Merge pull request kubeflow#2970 from lluunn/automated-cherry-pick-of…

…-#2968-upstream-v0.5-branch

Automated cherry pick of kubeflow#2968: set injectGcpCredentials to true by default Cherry pick of kubeflow#2968 on v0.5-branch. kubeflow#2968: set injectGcpCredentials to true by default

Merge pull request kubeflow#2970 from lluunn/automated-cherry-pick-of…

…-#2968-upstream-v0.5-branch

Automated cherry pick of kubeflow#2968: set injectGcpCredentials to true by default Cherry pick of kubeflow#2968 on v0.5-branch. kubeflow#2968: set injectGcpCredentials to true by default

Merge pull request kubeflow#2932 from lluunn/v05_cp_0403

Cherrypick for 0.5

Cherrypick for 0.5 (kubeflow#2908)

* add labels for notebook so that gcp credentials will be injected by webhook (kubeflow#2853)

* add labels for gcp cred

* kfctl set flag

* review comment

* review comment

* Update image tags for katib (kubeflow#2875)

* Style guide added, with supporting docs (kubeflow#2878)

* Let jupyter service account list/get/create secret (kubeflow#2887)

The jupyter service account can already create a deployment with a
secret mounted and read it that way, so its just more of a hassle to not
let it explicitly list secrets.

* add auth cmd (kubeflow#2873)

* remove profile from v0.5 (kubeflow#2877)

* remove profile from v0.5

* remove from test check list

* remove namespace selector (kubeflow#2893)

* delete endpoints (kubeflow#2838)

* add new function deleteEndpoints

* test import works

* fix

* revert test-worker tag

* get_endpoints_list

* break test

* add error msg

* message

* fix

* finish deleteEndpoints

* use member var

* not bail out if retry error

* add description

* A script to reset the basic auth secret. (kubeflow#2662)

* A script to reset the basic auth secret.

* Fix the script.

* Resolve comments.

* Update the Jupyter UI for 0.5 (kubeflow#2848)

* Add polling logic

* Get pvcs
* Return more info in Notebooks
* Update Makefile, update css/html
* Update Rok UI
* Minor fixes on html and flask debug mode
* Update add-notebook.js in Rok
* Add html for new notebook in ROK
* Make UI to poll the Backend for State

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Update the GCB script

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Organize file structure and refactor imports

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Sync Notebook name with Workspace Volume

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Use different warning if no DefStorClass

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Make backend REST

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Change Notebook status

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Misc changes

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Scroll on top to show the error

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Clean notebook table

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Add dialogs

If a user wants to delete a notebook, provide a confirm dialog

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Use term Namespace instead of Project

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Fix CSS rendering

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Make the polling exponential

The frontend will stop polling the backend exponentially.
If the state doesn't change, reduce the retries counter.

If the retries counter goes to zero, then double the time between
requests and reset the retries counter. Maximum poll time is 16 seconds.

If the state changes, or the user creates/deletes a Notebook or
selects a different namespace; reset the polling mechanism

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Handle no Status in Notebook CR

Signed-off-by: Kimonas Sotirchos <[email protected]>

* Create a minimal release process for kfctl. (kubeflow#2894)

* Create a minimal release process for kfctl.

* Instructions for building binaries and uploading to GitHub release page.
* We need to figure out if/how we cross-compile for linux.

Related to: kubeflow#2383

* * Add make rules to build and push the binary.
* tar and gzip the binary
* Add separate binaries for linux and Darwin.

* update notebook image (kubeflow#2895)

* fix patch string for admission webhook (kubeflow#2903)

* fix patch string

* cleanup

Merge pull request kubeflow#2892 from zabbasi/juser

Set user inside Notebook to jovyan

Merge pull request kubeflow#2868 from lluunn/v05_cp2

Cherrypick to 0.5

Merge pull request kubeflow#2849 from lluunn/v05_cp0328

Cherrypick to 0.5

cherrypick changes to V0.4 branch (kubeflow#2274)

* Add retry to loading the sample job (kubeflow#2199)

Loading the sample job has sporadically failures due to race condition creating the db foreign key. 
Retry on the job should succeed.

* cherrypick

* bump pipeline sdk version to 0.1.7 (kubeflow#2253)

* remove load sample job from pipeline config (kubeflow#2254)

Avoid using job to load samples into pipeline system. 
The K8s Job doesn't work well because it's immutable once created, ksonnet throw an error of "merging object with existing state failure" in case of upgrading the system.

* cherrypick

cherrypick changes to V0.4 branch (kubeflow#2274)

* Add retry to loading the sample job (kubeflow#2199)

Loading the sample job has sporadically failures due to race condition creating the db foreign key. 
Retry on the job should succeed.

* cherrypick

* bump pipeline sdk version to 0.1.7 (kubeflow#2253)

* remove load sample job from pipeline config (kubeflow#2254)

Avoid using job to load samples into pipeline system. 
The K8s Job doesn't work well because it's immutable once created, ksonnet throw an error of "merging object with existing state failure" in case of upgrading the system.

* cherrypick