Skip to content

Reduce concurrency bottlenecks #11

@vorce

Description

@vorce

At the moment an index is represented by a process. This is not great in terms of concurrency, a "hot" index (ie getting many requests) will receive a lot of messages and each will be handled one by one.

Perhaps a dynamic pool for each index? Or some sort of worker queue?

Metadata

Metadata

Assignees

No one assigned

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions