Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Use jemalloc #1321

Merged
merged 2 commits into from
Oct 22, 2021
Merged

Use jemalloc #1321

merged 2 commits into from
Oct 22, 2021

Conversation

olix0r
Copy link
Member

@olix0r olix0r commented Oct 22, 2021

In load tests, it appears that jemalloc is--perhaps unsurprisingly--more efficient
than the alternatives. Specifically, jemalloc uses less memory and is more responsive
to memory being released without substantially impacting latency or CPU usage.

This is only enabled on x86_64-unknown-linux-gnu, as this is the only platform
on which all tests pass, etc (per gnzlbg/jemallocator).


jemalloc vs system allocator vs mimalloc

image

@olix0r olix0r requested a review from a team October 22, 2021 01:07
Copy link
Contributor

@hawkw hawkw left a comment

Choose a reason for hiding this comment

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

cool!

@olix0r olix0r merged commit 0ac4f3f into main Oct 22, 2021
@olix0r olix0r deleted the ver/jemalloc branch October 22, 2021 03:17
olix0r added a commit to linkerd/linkerd2 that referenced this pull request Oct 22, 2021
This release changes the default allocator (on linux/x86_64) to be
`jemalloc`. In tests, jemalloc proves to use less memory without
noticeable impacts on CPU utilization or latency.

This change also includes dependency updates, including an update of
Rust to v1.56.0.

---

* build(deps): bump tower from 0.4.9 to 0.4.10 (linkerd/linkerd2-proxy#1320)
* Adopt Rust edition 2021 (linkerd/linkerd2-proxy#1295)
* Use jemalloc (linkerd/linkerd2-proxy#1321)
* Disable incremental compilation in CI (linkerd/linkerd2-proxy#1322)
* build(deps): bump h2 from 0.3.6 to 0.3.7 (linkerd/linkerd2-proxy#1324)
* build(deps): bump http-body from 0.4.3 to 0.4.4 (linkerd/linkerd2-proxy#1325)
* build(deps): bump hyper from 0.14.13 to 0.14.14 (linkerd/linkerd2-proxy#1326)
olix0r added a commit that referenced this pull request Oct 22, 2021
In load tests, it appears that jemalloc is--perhaps unsurprisingly--more efficient
than the alternatives. Specifically, jemalloc uses less memory and is more responsive
to memory being released without substantially impacting latency or CPU usage.

This is only enabled on `x86_64-unknown-linux-gnu`, as this is the only platform
on which all tests pass, etc (per [gnzlbg/jemallocator][readme]).

[readme]: https://github.com/gnzlbg/jemallocator/tree/c27a859e98e3cb790dc269773d9da71a1e918458#platform-support
alpeb pushed a commit to linkerd/linkerd2 that referenced this pull request Oct 25, 2021
This release changes the default allocator (on linux/x86_64) to be
`jemalloc`. In tests, jemalloc proves to use less memory without
noticeable impacts on CPU utilization or latency.

This change also includes dependency updates, including an update of
Rust to v1.56.0.

---

* build(deps): bump tower from 0.4.9 to 0.4.10 (linkerd/linkerd2-proxy#1320)
* Adopt Rust edition 2021 (linkerd/linkerd2-proxy#1295)
* Use jemalloc (linkerd/linkerd2-proxy#1321)
* Disable incremental compilation in CI (linkerd/linkerd2-proxy#1322)
* build(deps): bump h2 from 0.3.6 to 0.3.7 (linkerd/linkerd2-proxy#1324)
* build(deps): bump http-body from 0.4.3 to 0.4.4 (linkerd/linkerd2-proxy#1325)
* build(deps): bump hyper from 0.14.13 to 0.14.14 (linkerd/linkerd2-proxy#1326)
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