Skip to content

bug: exclude docker pull time from job execution timeouts & incorrect job state while pulling image #4795

@laciferin2024

Description

@laciferin2024

Bug Description

I set up a bacalhau node with default timeout to 30m. bacalhau job ran for 30m timed out. I debugged a saw that all the time when job was in progress, it was actually image being pulled. Ultimately when image took longer to pull.

A new status for jobs like SettingUp would be nice,

image

This status

   bacalhau logs -f 7c5a7a88-f181-4dca-87aa-54b6ae293633
error reading log output: error received from server: failed to open log stream for job 7c5a7a88-f181-4dca-87aa-54b6ae293633: failed to get log stream for execution: e-d9269b5c-4b03-4132-9222-659f21007b14. getting outputs for execution (e-d9269b5c-4b03-4132-9222-659f21007b14): execution not found
image

This false positive confirmation of job in Progress, crashes my workflow

Metadata

Metadata

Assignees

No one assigned

    Labels

    bacalhaurequest/newRequest: Indicates a new request that has been submitted and awaits initial triagetype/bugType: Something is not working as expected

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions