57 Star 258 Fork 126

GVPopenEuler / A-Tune

Create your Gitee Account
Explore and code with more than 6 million developers,Free private repositories !:)
Sign up
Clone or Download
Cancel
Notice: Creating folder will generate an empty file .keep, because not support in Git
Loading...
README.md

English | 简体中文

Introduction to A-Tune

A-Tune is an OS tuning engine powered by AI. A-Tune uses AI technologies to enable the OS to understand services, simplify IT system tuning, and maximize application performance.

I. A-Tune Installation

Supported OS: openEuler 20.03 LTS or later

Method 1 (applicable to common users): Use the default A-Tune of openEuler.

yum install -y atune

For openEuler 20.09 or later, atune-engine is needed.

yum install -y atune-engine

Method 2 (applicable to developers): Use the source code of the local repository for installation.

1. Install dependent system software packages.

yum install -y golang-bin python3 perf sysstat hwloc-gui

2. Install Python dependency packages.

2.1 Install dependency for the A-Tune service.

yum install -y python3-dict2xml python3-flask-restful python3-pandas python3-scikit-optimize python3-xgboost python3-pyyaml

Or

pip3 install dict2xml Flask-RESTful pandas scikit-optimize xgboost scikit-learn pyyaml

2.2 (Optional) Install dependency for the database.

If you have already installed the database application and want to store A-Tune collection and tuning data to the database, you must also install the following packages:

yum install -y python3-sqlalchemy python3-cryptography

Or

pip3 install sqlalchemy cryptography

To use the database, you should also select either of the following methods to install dependency for the database application.

Database Install Using yum Install Using pip
PostgreSQL yum install -y python3-psycopg2 pip3 install psycopg2

3. Download the source code.

git clone https://gitee.com/openeuler/A-Tune.git

4. Compile.

cd A-Tune
make

5. Install.

make collector-install
make install

II. Quick Guide

1. Configure the A-Tune service.

Modify the network and disk configuration in the atuned.cnf file.

You can run the following command to query the NIC that needs to be specified for data collection or optimization and change the value of the network configuration item in the /etc/atuned/atuned.cnf file to the specified NIC.

ip addr

You can run the following command to query the disk that needs to be specified for data collection or optimization and change the value of the disk configuration item in the /etc/atuned/atuned.cnf file to the specified disk.

fdisk -l | grep dev

2. Manage the A-Tune service.

Load and start the atuned and atune-engine services.

systemctl daemon-reload
systemctl start atuned
systemctl start atune-engine

Check the status of the atuned and atune-engine services.

systemctl status atuned
systemctl status atune-engine

3. Generate AI models.

You can save the newly collected data to the A-Tune/analysis/dataset directory and run the model generation tool to update the AI model in the A-Tune/analysis/models directory.

Format

python3 generate_models.py

Parameter Description

Parameter Description
--csv_path, -d Path for storing CSV files required for model training. The default directory is A-Tune/analysis/dataset.
--model_path, -m Path for storing the new models generated during training. The default path is A-Tune/analysis/models.
--select, -s Indicates whether to generate feature models. The default value is false.
--search, -g Indicates whether to enable parameter space search. The default value is false.

Example:

python3 generate_models.py

4. Run the atune-adm commands.

list command

This command is used to list the supported profiles as well as active profiles.

Format:

atune-adm list

Example:

atune-adm list

profile command

This command is used to manually activate the profile to make it in the active state.

Format:

atune-adm profile

Example: Activate the profile corresponding to the web-nginx-http-long-connection.

atune-adm profile web-nginx-http-long-connection

analysis command (online static tuning)

This command is used to collect real-time statistics from the system to identify and automatically optimize workload types.

Format:

atune-adm analysis [OPTIONS]

Example 1: Use the default model to identify applications and perform automatic tuning.

atune-adm analysis

Example 2: Use the user-defined model for recognition.

atune-adm analysis --model /usr/libexec/atuned/analysis/models/new-model.m

tuning command (offline dynamic tuning)

Use the specified project file to search the dynamic space for parameters and find the optimal solution under the current environment configuration.

Format:

atune-adm tuning [OPTIONS] <PROJECT_YAML>

Example: See the A-Tune offline tuning example. Each example has a corresponding README guide.

For details about other commands, see the atune-adm help information or A-Tune User Guide.

III. Web UI

A-Tune-UI is a web project based on A-Tune. Please check A-Tune-UI README for details.

IV. How to Contribute

We welcome new contributors to participate in the project, and we are happy to provide guidance for new contributors. Please sign CLA before contribution.

Mailing list

If you have any question, please contact A-Tune.

Routine meeting

The SIG meeting is hold at 10:00-12:00 AM on Friday every two weeks. Please send your issues to the A-Tune mailing list.

About

A-Tune is an OS tuning engine based on AI. expand collapse
Go and 4 more languages
MulanPSL-2.0
Cancel

Releases (4)

All

A-Tune

Contributors

All

Activities

Load More
can not load any more
Go
1
https://toscode.gitee.com/openeuler/A-Tune.git
git@toscode.gitee.com:openeuler/A-Tune.git
openeuler
A-Tune
A-Tune
master

Search

182229 41614e54 1850385 182230 7885ed45 1850385