Skip to content

Conversation

@wesm
Copy link
Member

@wesm wesm commented Sep 5, 2019

I propose we revert to the jemalloc 4.x stable branch where we were before until we understand how to get the performance we need for our workloads from 5.x. See upstream issue jemalloc/jemalloc#1621

@wesm
Copy link
Member Author

wesm commented Sep 5, 2019

@xhochy @pitrou I suggest that we merge this change until we can do appropriate profiling (and work with the jemalloc community) to bring 5.2.x performance inline with the stable-4 branch

Copy link
Member Author

@wesm wesm left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

+1. I will follow up with the jemalloc folks to do some profiling when I have some time

@wesm wesm changed the title ARROW-6417: [C++] Revert to jemalloc stable-4 until we understand 5.2.x performance issues ARROW-6478: [C++] Revert to jemalloc stable-4 until we understand 5.2.x performance issues Sep 6, 2019
@wesm wesm closed this in 53c5af0 Sep 6, 2019
@wesm wesm deleted the ARROW-6417-jemalloc branch September 6, 2019 16:04
Copy link
Member

@xhochy xhochy left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Being in mobile: does this also reintroduce vendoring in e.g. the conda package? (It should)

+1 from me for the change in general.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants