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

[gcp] deploy.sh should support rerunning deploy.sh when DM configs and ks app already exist #1284

Closed
jlewi opened this issue Jul 28, 2018 · 1 comment

Comments

@jlewi
Copy link
Contributor

jlewi commented Jul 28, 2018

We should support the following flow

  1. Set `KUBEFLOW_DEPLOY=false1
  2. Run deploy.sh
    • Creates the config directories but doesn't deploy
  3. User modifies the configs
  4. User runs deploy.sh
    • This deploys things

There are some minor bugs that prevent this flow from working with 0.2.2 and on master.

Some issues

  • deploy.sh will try to run ks init which will fail if app already exists
    * need to skip recreating the ksonnet app and modifying it if it already exists.
jlewi added a commit to jlewi/kubeflow that referenced this issue Jul 31, 2018
* Check if resources like secrets already exist and if they do don't
  recreate them.

* Prevent users from checking in their GCP secrets

  1. Store them in a directory with a .gitignore file
  2. Remove the secrets as soon as we are done storing them in the cluster.

* Allow GCFS_INSTANCE to be set to null; e.g.
export GCFS_INSTANCE=
If GCFS_INSTANCE is null don't set it to some default value; instead
we interpret this as the user doesn't want GCFS to be installed.

* Related to kubeflow#1282 can't create GCFS for legacy network.

* Related to kubeflow#1284 support rerunning deploy.sh
k8s-ci-robot pushed a commit that referenced this issue Jul 31, 2018
* Make it easier to rerun deploy.sh

* Check if resources like secrets already exist and if they do don't
  recreate them.

* Prevent users from checking in their GCP secrets

  1. Store them in a directory with a .gitignore file
  2. Remove the secrets as soon as we are done storing them in the cluster.

* Allow GCFS_INSTANCE to be set to null; e.g.
export GCFS_INSTANCE=
If GCFS_INSTANCE is null don't set it to some default value; instead
we interpret this as the user doesn't want GCFS to be installed.

* Related to #1282 can't create GCFS for legacy network.

* Related to #1284 support rerunning deploy.sh

* Fix bug.
@jlewi
Copy link
Contributor Author

jlewi commented Aug 10, 2018

Fixed by #1308

@jlewi jlewi closed this as completed Aug 10, 2018
saffaalvi pushed a commit to StatCan/kubeflow that referenced this issue Feb 11, 2021
* Make it easier to rerun deploy.sh

* Check if resources like secrets already exist and if they do don't
  recreate them.

* Prevent users from checking in their GCP secrets

  1. Store them in a directory with a .gitignore file
  2. Remove the secrets as soon as we are done storing them in the cluster.

* Allow GCFS_INSTANCE to be set to null; e.g.
export GCFS_INSTANCE=
If GCFS_INSTANCE is null don't set it to some default value; instead
we interpret this as the user doesn't want GCFS to be installed.

* Related to kubeflow#1282 can't create GCFS for legacy network.

* Related to kubeflow#1284 support rerunning deploy.sh

* Fix bug.
yanniszark pushed a commit to arrikto/kubeflow that referenced this issue Feb 15, 2021
* Add e2e test for from volume resume

* Resume experiment after completion

* Print controller logs

* Remove test prints

* Remove controller logs
surajkota pushed a commit to surajkota/kubeflow that referenced this issue Jun 13, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

No branches or pull requests

1 participant