The code for the NSDI'21 paper "BMC: Accelerating Memcached using Safe In-kernel Caching and Pre-stack Processing".

Overview

BMC

The code for the NSDI'21 paper "BMC: Accelerating Memcached using Safe In-kernel Caching and Pre-stack Processing".

BibTex entry available here.

BMC (BPF Memory Cache) is an in-kernel cache for memcached. It enables runtime, crash-safe extension of the Linux kernel to process specific memcached requests before the execution of the standard network stack. BMC does not require modification of neither the Linux kernel nor the memcached application. Running memcached with BMC improves throughput by up to 18x compared to the vanilla memcached application.

Requirements

Linux kernel v5.3 or higher is required to run BMC.

Other software dependencies are required to build BMC and Memcached-SR (see Building BMC and Building Memcached-SR).

Build instructions

Building BMC

BMC must be compiled with libbpf and other header files obtained from kernel sources. The project does not include the kernel sources, but the kernel-src-download.sh and kernel-src-prepare.sh scripts automate the download of the kernel sources and prepare them for the compilation of BMC.

These scripts require the following software to be installed:

gpg curl tar xz make gcc flex bison libssl-dev libelf-dev

The project uses llvm and clang version 9 to build BMC, but more recent versions might work as well:

llvm-9 clang-9

Note that libelf-dev is also required to build libbpf and BMC.

With the previous software installed, BMC can be built with the following:

$ ./kernel-src-download.sh
$ ./kernel-src-prepare.sh
$ cd bmc && make

After BMC has been successfully built, kernel sources can be removed by running the kernel-src-remove.sh script from the project root.

Building Memcached-SR

Memcached-SR is based on memcached v1.5.19. Building it requires the following software:

clang-9 (or gcc-9) automake libevent-dev

Either clang-9 or gcc-9 is required in order to compile memcached without linking issues. Depending on your distribution, you might also need to use the -Wno-deprecated-declarations compilation flag.

Memcached-SR can be built with the following:

$ cd memcached-sr 
$ ./autogen.sh
$ CC=clang-9 CFLAGS='-DREUSEPORT_OPT=1 -Wno-deprecated-declarations' ./configure && make

The memcached binary will be located in the memcached-sr directory.

Further instructions

TC egress hook

BMC doesn't attach the tx_filter eBPF program to the egress hook of TC, it needs to be attached manually.

To do so, you first need to make sure that the BPF is mounted, if it isn't you can mount it with the following command:

# mount -t bpf none /sys/fs/bpf/

Once BMC is running and the tx_filter program has been pinned to /sys/fs/bpf/bmc_tx_filter, you can attach it using the tc command line:

# tc qdisc add dev 
   
     clsact
   
# tc filter add dev 
   
     egress bpf object-pinned /sys/fs/bpf/bmc_tx_filter
   

After you are done using BMC, you can detach the program with these commands:

# tc filter del dev 
   
     egress
   
# tc qdisc del dev 
   
     clsact
   

And unpin the program with # rm /sys/fs/bpf/bmc_tx_filter

License

Files under the bmc directory are licensed under the GNU Lesser General Public License version 2.1.

Files under the memcached-sr directory are licensed under the BSD-3-Clause BSD license.

Cite this work

BibTex:

@inproceedings{265047,
	title        = {{BMC}: Accelerating Memcached using Safe In-kernel Caching and Pre-stack Processing},
	author       = {Yoann Ghigoff and Julien Sopena and Kahina Lazri and Antoine Blin and Gilles Muller},
	year         = 2021,
	month        = apr,
	booktitle    = {18th {USENIX} Symposium on Networked Systems Design and Implementation ({NSDI} 21)},
	publisher    = {{USENIX} Association},
	pages        = {487--501},
	isbn         = {978-1-939133-21-2},
	url          = {https://www.usenix.org/conference/nsdi21/presentation/ghigoff}
}
Owner
Orange
Open Source by Orange
Orange
Hardware accelerated, batchable and differentiable optimizers in JAX.

JAXopt Installation | Examples | References Hardware accelerated (GPU/TPU), batchable and differentiable optimizers in JAX. Installation JAXopt can be

Google 621 Jan 08, 2023
This repository is an implementation of paper : Improving the Training of Graph Neural Networks with Consistency Regularization

CRGNN Paper : Improving the Training of Graph Neural Networks with Consistency Regularization Environments Implementing environment: GeForce RTX™ 3090

THUDM 28 Dec 09, 2022
VLGrammar: Grounded Grammar Induction of Vision and Language

VLGrammar: Grounded Grammar Induction of Vision and Language

Yining Hong 27 Dec 23, 2022
This code provides a PyTorch implementation for OTTER (Optimal Transport distillation for Efficient zero-shot Recognition), as described in the paper.

Data Efficient Language-Supervised Zero-Shot Recognition with Optimal Transport Distillation This repository contains PyTorch evaluation code, trainin

Meta Research 45 Dec 20, 2022
Code and data for the EMNLP 2021 paper "Just Say No: Analyzing the Stance of Neural Dialogue Generation in Offensive Contexts". Coming soon!

ToxiChat Code and data for the EMNLP 2021 paper "Just Say No: Analyzing the Stance of Neural Dialogue Generation in Offensive Contexts". Install depen

Ashutosh Baheti 11 Jan 01, 2023
Deep Reinforcement Learning with pytorch & visdom

Deep Reinforcement Learning with pytorch & visdom Sample testings of trained agents (DQN on Breakout, A3C on Pong, DoubleDQN on CartPole, continuous A

Jingwei Zhang 783 Jan 04, 2023
Cosine Annealing With Warmup

CosineAnnealingWithWarmup Formulation The learning rate is annealed using a cosine schedule over the course of learning of n_total total steps with an

zhuyun 4 Apr 18, 2022
Locally Enhanced Self-Attention: Rethinking Self-Attention as Local and Context Terms

LESA Introduction This repository contains the official implementation of Locally Enhanced Self-Attention: Rethinking Self-Attention as Local and Cont

Chenglin Yang 20 Dec 31, 2021
RL algorithm PPO and IRL algorithm AIRL written with Tensorflow.

RL algorithm PPO and IRL algorithm AIRL written with Tensorflow. They have a parallel sampling feature in order to increase computation speed (especially in high-performance computing (HPC)).

Fangjian Li 3 Dec 28, 2021
WSDM2022 "A Simple but Effective Bidirectional Extraction Framework for Relational Triple Extraction"

BiRTE WSDM2022 "A Simple but Effective Bidirectional Extraction Framework for Relational Triple Extraction" Requirements The main requirements are: py

9 Dec 27, 2022
This repository holds code and data for our PETS'22 article 'From "Onion Not Found" to Guard Discovery'.

From "Onion Not Found" to Guard Discovery (PETS'22) This repository holds the code and data for our PETS'22 paper titled 'From "Onion Not Found" to Gu

Lennart Oldenburg 3 May 04, 2022
Official implementation of the paper "Lightweight Deep CNN for Natural Image Matting via Similarity Preserving Knowledge Distillation"

Lightweight-Deep-CNN-for-Natural-Image-Matting-via-Similarity-Preserving-Knowledge-Distillation Introduction Accepted at IEEE Signal Processing Letter

DongGeun-Yoon 19 Jun 07, 2022
LOFO (Leave One Feature Out) Importance calculates the importances of a set of features based on a metric of choice,

LOFO (Leave One Feature Out) Importance calculates the importances of a set of features based on a metric of choice, for a model of choice, by iteratively removing each feature from the set, and eval

Ahmet Erdem 691 Dec 23, 2022
Natural Intelligence is still a pretty good idea.

Human Learn Machine Learning models should play by the rules, literally. Project Goal Back in the old days, it was common to write rule-based systems.

vincent d warmerdam 641 Dec 26, 2022
Mind the Trade-off: Debiasing NLU Models without Degrading the In-distribution Performance

Models for natural language understanding (NLU) tasks often rely on the idiosyncratic biases of the dataset, which make them brittle against test cases outside the training distribution.

Ubiquitous Knowledge Processing Lab 22 Jan 02, 2023
Hydra Lightning Template for Structured Configs

Hydra Lightning Template for Structured Configs Template for creating projects with pytorch-lightning and hydra. How to use this template? Create your

Model-driven Machine Learning 4 Jul 19, 2022
Official repo for QHack—the quantum machine learning hackathon

Note: This repository has been frozen while we consider the submissions for the QHack Open Hackathon. We hope you enjoyed the event! Welcome to QHack,

Xanadu 118 Jan 05, 2023
Pytorch implementation of our method for high-resolution (e.g. 2048x1024) photorealistic video-to-video translation.

vid2vid Project | YouTube(short) | YouTube(full) | arXiv | Paper(full) Pytorch implementation for high-resolution (e.g., 2048x1024) photorealistic vid

NVIDIA Corporation 8.1k Jan 01, 2023
PyTorch implementation of paper "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", CVPR 2021

Neural Scene Flow Fields PyTorch implementation of paper "Neural Scene Flow Fields for Space-Time View Synthesis of Dynamic Scenes", CVPR 20

Zhengqi Li 585 Jan 04, 2023
A curated list of awesome game datasets, and tools to artificial intelligence in games

🎮 Awesome Game Datasets In computer science, Artificial Intelligence (AI) is intelligence demonstrated by machines. Its definition, AI research as th

Leonardo Mauro 454 Jan 03, 2023