Skip to content

Conversation

@Dhanush010
Copy link
Contributor

When using @Retry(times=0) with JobSet through Argo Workflows, the workflow fails because {{retries}} is added to the jobset-name even though retryStrategy is not specified in the template.

The issue was that the code checked for the presence of the @Retry decorator instead of checking if retries are actually enabled (total_retries > 0). When total_retries = 0, retry_strategy() doesn't add retryStrategy, so {{retries}} is not available in Argo Workflows.

Fix by computing total_retries and only adding retry-related parameters (including {{retries}} in the jobset-name) when total_retries > 0.

Fixes #2632

When using @Retry(times=0) with JobSet through Argo Workflows, the
workflow fails because {{retries}} is added to the jobset-name even
though retryStrategy is not specified in the template.

The issue was that the code checked for the presence of the @Retry
decorator instead of checking if retries are actually enabled
(total_retries > 0). When total_retries = 0, retry_strategy() doesn't
add retryStrategy, so {{retries}} is not available in Argo Workflows.

Fix by computing total_retries and only adding retry-related parameters
(including {{retries}} in the jobset-name) when total_retries > 0.

Fixes Netflix#2632
@savingoyal savingoyal requested a review from valayDave November 18, 2025 19:06
@Dhanush010
Copy link
Contributor Author

Hi,
I've fixed the pre-commit formatting issue. The black formatter required the max() call to be split across multiple lines. Applied the formatting fix and pushed the changes - the pre-commit check should pass now.

@Dhanush010 Dhanush010 requested a review from valayDave November 20, 2025 19:13
@Dhanush010
Copy link
Contributor Author

Hi @valayDave,
The test failure is due to a transient infrastructure issue, not related to the code changes. The error shows:
Error: failed to fetch https://github.com/argoproj/argo-helm/releases/download/argo-events-2.4.8/argo-events-2.4.8.tgz : 503 Service Unavailable

This occurred during the "Bring up the environment" step when trying to download the argo-events Helm chart. The failure happened before any actual tests ran. This is a GitHub release download service issue (503 error), not a problem with the code changes.

Could someone please re-run the test? The other 31 checks passed successfully.

@savingoyal savingoyal enabled auto-merge (squash) November 21, 2025 19:48
@savingoyal savingoyal merged commit a881605 into Netflix:master Nov 21, 2025
32 checks passed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

Issue when using Jobset with @retry(times=0)

3 participants