Skip to content

Performance regressions for deeply-nested terraform modules #7524

@patrickvinograd

Description

@patrickvinograd

At various points in time we've observed serious performance regressions while scanning our terraform projects, notably those that have deeply nested modules and lots of use of count/for_each. I have notes from my teammate who investigated, saying:

3.2.34 took runs from 15->30m, then 3.2.119 took them to 4h+

At the time that teammate developed a patch that avoids some object allocation inside inner loops, and that brought performance back to acceptable levels. We've been carrying on applying that patch when we install checkov, but it recently broke and rather than patch the patch, we thought we'd try to upstream the fix.

I already created a PR but am filing an issue to ensure that the issue gets the proper visibility.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions