Skip to content

Conversation

@engelke
Copy link
Contributor

@engelke engelke commented Apr 4, 2023

Description

Does not use different projects for different versions of Python any more.

@product-auto-label product-auto-label bot added samples Issues that are directly related to samples. api: appengine Issues related to the App Engine Admin API API. api: asset api: batch Issues related to the Batch API. api: cloudkms Issues related to the Cloud Key Management Service API. api: compute Issues related to the Compute Engine API. api: dataproc Issues related to the Dataproc API. api: dialogflow Issues related to the Dialogflow API. api: iam Issues related to the Identity and Access Management API. api: jobs Issues related to the Cloud Talent Solution Job Search API. api: monitoring Issues related to the Cloud Monitoring API. api: recaptchaenterprise Issues related to the reCAPTCHA API. api: security-privateca Issues related to the Certificate Authority Service API. api: healhcare labels Apr 4, 2023
@nicain nicain requested a review from parthea April 4, 2023 21:13
# to use your own Cloud project.
# "gcloud_project_env": "GOOGLE_CLOUD_PROJECT",
"gcloud_project_env": "BUILD_SPECIFIC_GCLOUD_PROJECT",
"gcloud_project_env": "GOOGLE_CLOUD_PROJECT",
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This change, combined with #9489 and upgrading python-docs-samples-tests to Enterprise, should permanently resolve the Dialogflow-related error

@engelke engelke added the kokoro:force-run Add this label to force Kokoro to re-run the tests. label Apr 4, 2023
@kokoro-team kokoro-team removed the kokoro:force-run Add this label to force Kokoro to re-run the tests. label Apr 5, 2023
Copy link
Contributor

@m-strzelczyk m-strzelczyk left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I'm worried that we might face some new problems after squeezing all the tests into single project. Things like quota or tests interfering with each other.

I assume you have good reasons to change this for the samples you're investigating, so no veto from me in general. I'd only want to keep Batch tests in separate projects, as explained in my comment there.

# to use your own Cloud project.
# "gcloud_project_env": "GOOGLE_CLOUD_PROJECT",
"gcloud_project_env": "BUILD_SPECIFIC_GCLOUD_PROJECT",
"gcloud_project_env": "GOOGLE_CLOUD_PROJECT",
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I believe that this is build specific for the VM quota issue as well - dataproc instances count towards the VM quota

# to use your own Cloud project.
# "gcloud_project_env": "GOOGLE_CLOUD_PROJECT",
"gcloud_project_env": "BUILD_SPECIFIC_GCLOUD_PROJECT",
"gcloud_project_env": "GOOGLE_CLOUD_PROJECT",
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I believe that this is build specific for the VM quota issue as well - dataproc instances count towards the VM quota

# to use your own Cloud project.
# "gcloud_project_env": "GOOGLE_CLOUD_PROJECT",
"gcloud_project_env": "BUILD_SPECIFIC_GCLOUD_PROJECT",
"gcloud_project_env": "GOOGLE_CLOUD_PROJECT",
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I believe that this is build specific for the VM quota issue as well - dataproc instances count towards the VM quota

@engelke
Copy link
Contributor Author

engelke commented Apr 5, 2023

Small variations between the different test projects have led to a lot of very hard to debug flaky errors. Using a single project will eliminate that problem, while making quota problems more likely. We have managed to resolve most of those problems (future PRs will be coming soon after this one is merged), except for:

  • GAE send mail per minute (resolved) and per day quotas. The 100 emails per day quota is a problem when three samples are run in five versions of Python.

  • Some compute/client-library tests that currently try a large list of variations. The tests usually don't cause quota issues, but some remain sometimes. Perhaps just testing a smaller representative number of variations will suffice.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

api: appengine Issues related to the App Engine Admin API API. api: asset api: batch Issues related to the Batch API. api: cloudkms Issues related to the Cloud Key Management Service API. api: compute Issues related to the Compute Engine API. api: dataproc Issues related to the Dataproc API. api: dialogflow Issues related to the Dialogflow API. api: healhcare api: iam Issues related to the Identity and Access Management API. api: jobs Issues related to the Cloud Talent Solution Job Search API. api: monitoring Issues related to the Cloud Monitoring API. api: recaptchaenterprise Issues related to the reCAPTCHA API. api: security-privateca Issues related to the Certificate Authority Service API. samples Issues that are directly related to samples.

Projects

None yet

Development

Successfully merging this pull request may close these issues.

6 participants