Skip to content

enumerating release jobset names is single-threaded and slow #269357

@ghost

Description

The only known way to enumerate the names of all packages/jobs in a release is the one used by ofborg for the eval-checks:

This is obscenely slow and requires an absurd amount of memory. It is entirely single-threaded with zero parallelism. It peaks at 60gbytes of memory and takes 41 minutes on a 3ghz cpu.

This accounts for the vast majority of the cycle time for our CI.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions