Skip to content

Add async store and pipelining code#4

Merged
YaoJiayi merged 3 commits intoLMCache:mainfrom
YaoJiayi:main
Jun 17, 2024
Merged

Add async store and pipelining code#4
YaoJiayi merged 3 commits intoLMCache:mainfrom
YaoJiayi:main

Conversation

@YaoJiayi
Copy link
Copy Markdown
Collaborator

No description provided.

@YaoJiayi YaoJiayi merged commit eff09fd into LMCache:main Jun 17, 2024
@ApostaC
Copy link
Copy Markdown
Contributor

ApostaC commented Jun 17, 2024

A general comment on the design choices in coding:
If we want to support two different implementations of the same "logical operation", please consider implementing 2 different classes instead of writing new functions in the same class.
For example, when implementing pipeline optimization for get, it would be better to have a new class derived from LMCBackendInterface and override the batched_get function instead of adding a new function called batched_get_pipeline in the existing class.

This has at least 2 benefits:

  • you don’t need to change the processing logic of the callers (so the callers will call the same function, instead of calling a new function named batched_get_pipeline
  • If there is a configuration item that controls the use of the pipeline or not, you only need to check the configuration once during the initialization. Otherwise, you need to check the configuration during the runtime, which makes the code less readable.

Note that this is less about the runtime performance but more related to the readability and extendability of the code.

@ApostaC
Copy link
Copy Markdown
Contributor

ApostaC commented Jun 17, 2024

Also, we need to have the updated Dockerfile and the deploy instructions (now it does not include installation of torchac_cuda)

robertgshaw2-redhat added a commit to robertgshaw2-redhat/LMCache that referenced this pull request Apr 21, 2025
* [Add] optimized NIXL backend (WIP)

* [Add] better async primitives

* update cache engine log message

Signed-off-by: ApostaC <yihua98@uchicago.edu>

* [stash] debug works

Signed-off-by: ApostaC <yihua98@uchicago.edu>

* [Fix] performance and correctness bugs

Signed-off-by: ApostaC <yihua98@uchicago.edu>

* fix format checker issues

Signed-off-by: ApostaC <yihua98@uchicago.edu>

* [fix] isort errors

Signed-off-by: ApostaC <yihua98@uchicago.edu>

* fix format

* fix format again

* disable debug logs

* update

Signed-off-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>

* updated

Signed-off-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>

* merge conflict

Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>

* updated

Signed-off-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>

* updated

Signed-off-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>

* updated

Signed-off-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>

---------

Signed-off-by: ApostaC <yihua98@uchicago.edu>
Signed-off-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>
Signed-off-by: Tyler Michael Smith <tyler@neuralmagic.com>
Co-authored-by: ApostaC <yihua98@uchicago.edu>
Co-authored-by: rshaw@neuralmagic.com <robertgshaw2@gmail.com>
NumberWan pushed a commit to NumberWan/LMCache that referenced this pull request Sep 10, 2025
KevinCheung2259 pushed a commit to KevinCheung2259/LMCache that referenced this pull request Nov 5, 2025
Add async store and pipelining code
DongDongJu referenced this pull request in DongDongJu/LMCache Feb 22, 2026
Add async store and pipelining code
yoo-kumaneko pushed a commit to yoo-kumaneko/LMCache that referenced this pull request Mar 23, 2026
…request the current log belongs to (LMCache#4)

LMCache#2812

Signed-off-by: baoloongmao <baoloongmao@tencent.com>
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