Skip to content

GUI/API scalability for huge workflows #330

@LuiggiTenorioK

Description

@LuiggiTenorioK

For some time, inside the Climate DT VM, there have been many experiments that handle more than 100,000 jobs. This is a problem because many operators cannot use the API/GUI to have insights about these huge experiments.

To address that, we need to redesign some GUI pages to not retrieve and show all the information at once, and rather show small groups of information that are searchable. For that, we need to create some new API endpoints to serve experiment data in smaller chunks and use strategies like pagination or maybe caching some request responses.

To ensure this strategy actually improves the user experience, it is necessary that BSC-ES/autosubmit#2391 is ready and becomes the new way to store the job data. This is because the current job data can only be loaded as a whole and not in a segmented way.

For the first phase, the Table and Quick View of the GUI are targeted in this redesign since they seem to be the ones operators use most when going to huge workflows, and are less complex to modify.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    Status

    Todo

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions