[Misc] Improve nixl perf in lmcache mp by YaoJiayi · Pull Request #2711 · LMCache/LMCache

YaoJiayi · 2026-03-07T01:17:23Z

What this PR does / why we need it:
Specify file size in nixl backend.
Special notes for your reviewers:

If applicable:

this PR contains user facing changes - docs added
this PR contains unit tests

Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn>

gemini-code-assist · 2026-03-07T01:17:40Z

Summary of Changes

Hello, I'm Gemini Code Assist¹! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed!

This pull request enhances the performance of Nixl within LMCache's multiprocess environment by optimizing file-based storage management. It introduces a mechanism to consolidate multiple L1 memory pages into larger storage files, which streamlines I/O operations and reduces system resource overhead. Additionally, a minor adjustment was made to prevent busy-waiting in a critical lookup path, contributing to overall system efficiency.

Highlights

Nixl File Storage Optimization: Introduced a 'file_size' parameter for file-based Nixl backends, allowing multiple L1 pages to be stored within a single file, thereby reducing the number of file descriptors and improving I/O efficiency.
Configuration Update: Updated the configuration help text to include the new 'file_size' parameter for Nixl backend parameters, providing clearer guidance for users.
Busy-Waiting Prevention: Added a small 'time.sleep' delay in the 'lookup' method's polling loop to prevent busy-waiting and reduce CPU consumption during status checks.

🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console.

Changelog

lmcache/v1/distributed/l2_adapters/config.py
- Updated the help string for Nixl backend parameters to include the new 'file_size' option.
lmcache/v1/distributed/l2_adapters/nixl_store_l2_adapter.py
- Refactored the initialization of 'NixlObjPool' to be conditional on the backend type.
- Introduced 'file_size' parameter for file-based storage handlers, allowing a single file to contain multiple pages.
- Modified 'init_storage_handlers_file' to accept 'num_files' and 'file_size', and to correctly register memory and transfer descriptors for multi-page files.
- Added validation to ensure 'file_size' is a multiple of 'page_size'.
lmcache/v1/multiprocess/server.py
- Added a 'time.sleep' call in the 'lookup' method's polling loop to prevent busy-waiting.

Activity

No specific activity (comments, reviews, or progress updates) has been recorded for this pull request yet.

Using Gemini Code Assist

The full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips.

Invoking Gemini

You can request assistance from Gemini at any point by creating a comment using either /gemini <command> or @gemini-code-assist <command>. Below is a summary of the supported commands on the current page.

Feature	Command	Description
Code Review	`/gemini review`	Performs a code review for the current pull request in its current state.
Pull Request Summary	`/gemini summary`	Provides a summary of the current pull request in its current state.
Comment	@gemini-code-assist	Responds in comments when explicitly tagged, both in pull request comments and review comments.
Help	`/gemini help`	Displays a list of available commands.

Customization

To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a .gemini/ folder in the base of the repository. Detailed instructions can be found here.

Limitations & Feedback

Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here.

You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension.

Review the Privacy Notices, Generative AI Prohibited Use Policy, Terms of Service, and learn how to configure Gemini Code Assist in GitHub here. Gemini can make mistakes, so double check it and use code with caution. ↩

gemini-code-assist

Code Review

This pull request introduces a performance improvement for the nixl storage backend by allowing multiple pages to be grouped into a single file, which reduces the number of open file descriptors. The changes are primarily in nixl_store_l2_adapter.py to support a file_size parameter. Additionally, a time.sleep() is added to a busy-wait loop in server.py to reduce CPU consumption.

The overall approach is solid, but I've identified a critical bug in the new implementation within nixl_store_l2_adapter.py where an incorrect variable is passed to a function, which would likely cause transfers to fail. Please see the specific comment for details and a suggested fix.

Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn>

ApostaC

LGTM!

ApostaC

one minor comment

ApostaC · 2026-03-07T03:52:04Z

            if found_count is not None:
                break
+
+            time.sleep(0.001)


Let's don't have this here but have #2710 merged. That will the problem here

KuntaiDu

lgtm!

* add file size config Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> * fix minor Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> * revert some logs Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> --------- Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn>

* add file size config Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> * fix minor Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> * revert some logs Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> --------- Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> Signed-off-by: shaoxiawjc <wjc2800@163.com>

* add file size config Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> * fix minor Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> * revert some logs Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> --------- Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> Signed-off-by: Aaron Wu <aaron.wu@dell.com>

* add file size config Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> * fix minor Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> * revert some logs Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn> --------- Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn>

add file size config

e5ef20b

Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn>

gemini-code-assist Bot reviewed Mar 7, 2026

View reviewed changes

YaoJiayi added 2 commits March 7, 2026 01:31

fix minor

11f9195

Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn>

revert some logs

61767b9

Signed-off-by: YaoJiayi <120040070@link.cuhk.edu.cn>

ApostaC approved these changes Mar 7, 2026

View reviewed changes

ApostaC added the full Run comprehensive tests on this PR label Mar 7, 2026

ApostaC reviewed Mar 7, 2026

View reviewed changes

ApostaC requested a review from KuntaiDu March 7, 2026 03:59

KuntaiDu approved these changes Mar 7, 2026

View reviewed changes

YaoJiayi merged commit 284176a into dev Mar 7, 2026
23 of 28 checks passed

YaoJiayi deleted the localdev/nixl-v2 branch March 10, 2026 17:36

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[Misc] Improve nixl perf in lmcache mp#2711

[Misc] Improve nixl perf in lmcache mp#2711
YaoJiayi merged 3 commits intodevfrom
localdev/nixl-v2

YaoJiayi commented Mar 7, 2026 •

edited

Loading

Uh oh!

gemini-code-assist Bot commented Mar 7, 2026

Uh oh!

gemini-code-assist Bot left a comment

Uh oh!

ApostaC left a comment

Uh oh!

ApostaC left a comment

Uh oh!

ApostaC Mar 7, 2026

Uh oh!

KuntaiDu left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

YaoJiayi commented Mar 7, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

gemini-code-assist Bot commented Mar 7, 2026

Summary of Changes

Highlights

Footnotes

Uh oh!

gemini-code-assist Bot left a comment

Choose a reason for hiding this comment

Code Review

Uh oh!

ApostaC left a comment

Choose a reason for hiding this comment

Uh oh!

ApostaC left a comment

Choose a reason for hiding this comment

Uh oh!

ApostaC Mar 7, 2026

Choose a reason for hiding this comment

Uh oh!

KuntaiDu left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

YaoJiayi commented Mar 7, 2026 •

edited

Loading