Skip to content

Optimize tuple deepcopy #274

@DifferentialOrange

Description

@DifferentialOrange

After #268, input tuple is copied so we won't damage the original one. Performance has dropped by 5%.

Raw thoughts on possible optimizations:

  • Skip copying on crud.insert_object().
  • Pass the bucket_id separately to storage and insert it into the tuple on the storage (take care to router/storage rolling code update).

I don't want to change it within this pull request, but rather to think around later.

Originally posted by @Totktonada in #268 (comment)

Metadata

Metadata

Assignees

No one assigned

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions