When updating cloud run jobs image, job task-timeout is reset to default value.

[skandertajine]

Issue Description:

When updating a Cloud Run job's image version using the gcloud run command with the gcr.io/google.com/cloudsdktool/cloud-sdk:alpine, the task-timeout parameter of the cloud run job is being reset to 10 minutes.

Steps to Reproduce:

Use the following command to update the Cloud Run job's image:

gcloud run jobs update "${CLOUDRUN_INSTANCE_NAME}"
--image "${DOCKER_ROOT_IMAGE}"
--region "${CLOUDRUN_REGION}"
--project "${GCP_PROJECT_ID}"

Expected Behavior:
The timeout should retain its configured value during the image update process.

Actual Behavior:
The timeout is reset to 10 minutes.

Environment:

• gcr.io/google.com/cloudsdktool/cloud-sdk:slim
• gcloud GA version
• gitlab ci for running the update image step

Additional Information:
This behavior is causing issues in scenarios where a different timeout is configured, leading to unexpected behavior during the image update process.

Workaround
used gcr.io/google.com/cloudsdktool/cloud-sdk:alpine to maintain the task-timeout value.

Note:
Please let me know if any additional information or logs are required to investigate and resolve this issue.

[bobidle]

The slim image was recently upgraded to Debian Bookworm.

Are only the latest images based on Debian Bookworm affected?

• gcr.io/google.com/cloudsdktool/cloud-sdk:458.0.0-slim
• gcr.io/google.com/cloudsdktool/cloud-sdk:457.0.0-slim

Or also images based on Debian Bullseye

• gcr.io/google.com/cloudsdktool/cloud-sdk:456.0.0-slim
• or older