`Qlib-Server` Documentation¶

Qlib-Server is the assorted server system for Qlib, which utilizes Qlib for basic calculations and provides extensive server system and cache mechanism. With Qlib-Server, the data provided for Qlib can be managed in a centralized manner.

Document Structure¶

`Qlib-Server`: Quant Library Data Server¶

Introduction¶

Qlib-Server is the assorted server system for Qlib, which utilizes Qlib for basic calculations and provides extensive server system and cache mechanism. With Qlib-Server, the data provided for Qlib can be managed in a centralized manner.

Framework¶

The Client/Server framework of Qlib is based on WebSocket considering its capability of bidirectional communication between client and server in async mode.

Qlib-Server is based on Flask, which is a micro-framework for Python and here Flask-SocketIO is used for websocket connection.

Qlib-Server provides the following procedures:

Listening to incoming request from client¶

The clients will propose several types of requests to server. The server will parse the requests, collect the identical requests from different clients, record their session-ids, and submit these parsed tasks to a pipe. Qlib use RabbitMQ as this pipe. The tasks will be published to a channel task_queue.

RequestListener is used for this function:

After receiving these requests, the server will check whether different clients are asking for the same data. If so, to prevent repeated generation of data or repeated generation of cache files, the server will use Redis to maintain the session-ids of those clients. These session-ids will be deleted once this task is finished. To avoid IO conflicts, Redis_Lock is imported to make sure no tasks in redis will be read and written at the same time.

Responding clients with data¶

The server consumes the result from message_queue and get the session-ids of the clients requring this result. Then it responds to these clients with results.

RequestResponder is used for this method.

The two class above is combined as RequestHandler, which is responsible for communicating with clients.

Accepting tasks from RabbitMQ and processing data¶

The server will automatically collect tasks from RabbitMQ and process the relevant data. RabbitMQ provides a mechanism that when the server consumes a task, a callback function is triggered. The data processing procedure is implemented within these callbacks and it currently supports the three types of tasks corresponding to above:

Calendar
Instruments
Features

DataProcessor is used for this function.

The server will use qlib.data.Provider to process the data. RabbitMQ also provides a mechanism that can make sure all tasks is succesfully consumed and completed by consumers. This requires the consumer call ch.basic_ack(delivery_tag=method.delivery_tag) after succesfully processing the data. If the task is not acked, it will return to the pipe and wait for another consuming.

Once the task is finished, a result (could be data or uri) will be published to another channel message_queue.

`Qlib-Server` Deployment¶

Introduction¶

To build a Qlib-Server, user can choose:

One-click Deployment of Qlib-Server
Step-by-step Deployment of Qlib-Server

One-click Deployment¶

One-click deployment of Qlib-Server is supported, users can choose either of the following two methods for one-click deployment:

Deployment with docker-compose
Deployment in Azure

One-click Deployment with `docker-compose`¶

Deploy Qlib-Server with docker-compose according to the following processes:

Install docker, please refer to Docker Installation.
Install docker-compose, please refer to Docker-compose Installation.

Run the following command to deploy Qlib-Server:

git clone https://github.com/microsoft/qlib-server
cd qlib-server
sudo docker-compose -f docker_support/docker-compose.yaml --env-file docker_support/docker-compose.env build
sudo docker-compose -f docker_support/docker-compose.yaml --env-file docker_support/docker-compose.env up -d
# Use the following command to track the log
sudo docker-compose -f docker_support/docker-compose.yaml logs -f

One-click Deployment in `Azure`¶

Note

Users need to have an Azure account to deploy Qlib-Server in Azure.

Deploy Qlib-Server in Azure according to the following processes:

Install azure-cli, please refer to install-azure-cli

Add the Azure account to the configuration file azure_conf.yaml

sub_id: Your Subscription ID
username: azure user name
password: azure password
# The resource group where the VM is located
resource_group: Resource group name

Execute the deployment script

Run the following command:

To know more about parameters, please run the following command:

Step-by-step Deployment¶

Users can deploy Qlib-Server step by step, which has the following processes:

Build RabbitMQ
Build Redis
Build NFS
Build Qlib-Server

Build `RabbitMQ`¶

RabbitMQ is a general task queue that enables qlib-server to separate request handling process and data generating process.

Note

Users need not to build RabbitMQ instance on the same server as Qlib-Server.

Build RabbitMQ according to the following processes:

Import RabbitMQ signing key on your system:

echo 'deb http://www.rabbitmq.com/debian/ testing main' | sudo tee /etc/apt/sources.list.d/rabbitmq.list
wget -O- https://www.rabbitmq.com/rabbitmq-release-signing-key.asc | sudo apt-key add -

Update apt cache and install RabbitMQ server on your system:

sudo apt-get update
sudo apt-get install rabbitmq-server

Enable the RabbitMQ service and start it.

# Using Init –
sudo update-rc.d rabbitmq-server defaults
sudo service rabbitmq-server start
sudo service rabbitmq-server stop

# Using Systemctl -
sudo systemctl enable rabbitmq-server
sudo systemctl start rabbitmq-server
sudo systemctl stop rabbitmq-server

Create admin user in RabbitMQBy

By default RabbitMQBy creates a username guest with password guest. Users can also create admin user in RabbitMQ:

sudo rabbitmqctl add_user admin <your password>
sudo rabbitmqctl set_user_tags admin administrator
sudo rabbitmqctl set_permissions -p / admin ".*" ".*" ".*"

Enable web management console
RabbitMQ also provides and web management console for managing the entire RabbitMQ. To enable web management console run following command. The web management console helps users with managing RabbitMQ server.
sudo rabbitmq-plugins enable rabbitmq_management
Visit <your rabbitmq host>:15672 to manage your queue. Keep in mind your rabbitmq host and credentials. It will be used in qlib-server config.

Build `Redis`¶

Qlib-Server needs redis to store and read some meta info as well as thread lock.

Note

Users need not to build redis instance on the same server as Qlib-Server.

Build redis according to the following processes:

Download the latest version of redis and install

mkdir ~/redis
cd ~/redis
wget http://download.redis.io/releases/redis-5.0.4.tar.gz
tar -zxvf redis-5.0.4.tar.gz
cd redis-5.0.4
sudo make && make install

Start redis service
/usr/local/bin/redis-server
The default port of redis is 6379. Keep in mind your redis host and port. It will be used in qlib-server config.

Build `NFS`¶

Before starting Qlib-Server, it’s necessary to make sure the cache file directories are mounted (or at least ready to be mounted) to clients by configuring nfs service.

Build NFS according to the following processes:

Install NFS service:
sudo apt-get install nfs-kernel-server
Check if the nfs port is open:
netstat -tl
Note

By seeing tcp 0 0 *:nfs *:* LISTEN, the nfs port is ready for listening. Restart the service to ensure it can be used:
sudo /etc/init.d/nfs-kernel-server restart
Modify /etc/exports to give the directories ability to be mounted. To find out how the keywords like rw work and change them, please refer to nfs documents.

sudo echo '<your data directory> *(rw,sync,no_subtree_check,no_root_squash)'>>/etc/exports

Use showmount to view the exported directories.

Build `Qlib-Server`¶

Users can choose one of the following two methods to build Qlib-Server:

Build with Source Code
Build with Dockerfile

Build with Source Code¶

Build Qlib-Server with source code according to the following processes:

Enter the Qlib-Server directory and run python setup.py install.
Modify the config.yaml according to users’ needs and configs.

Start using Qlib-server by running:

cp config_template.yaml config.yaml
edit config.yaml  # Please edit the server config.
python main.py -c config.yaml

Warning

Rabbitmq and redis configurations cannot be shared among multiple qlib-server instances

Eg:

In config_1.yaml, redis_db:1 task_queue: 'task_queue_1' √
In config_2.yaml, redis_db:2 task_queue: 'task_queue_2' √
---------------------------------------------------------
In config_1.yaml, redis_db:1 task_queue: 'task_queue_1' ×
In config_2.yaml, redis_db:1 task_queue: 'task_queue_1' ×

Note

The content of config.yaml is as follows

provider_uri

Qlib data directory
flask_server

Flask server host/ip, can be 0.0.0.0 or private ip
flask_port

Data service port, with which the client port must be consistent to access server
queue_host

RabbitMQ server ip/host
queue_user

RabbitMQ user name
queue_pwd

RabbitMQ password
task_queue

Task queue of Qlib-Server, if rabbitmq serves multiple Qlib-Server s, this value cannot be repeated
message_queue

Message queue of Qlib-Server, if rabbitmq serves multiple Qlib-Server s, this value cannot be repeated
redis_host

Redis server host/ip
redis_port

Redis server port
redis_task_db

Redis database name
auto_update

Currently, this parameter is not used
update_time

Currently, this parameter is not used
client_version

The version of Qlib must be newer than client_version to access the Qlib-Server
server_version

The version of Qlib must be newer than server_version to install or run Qlib-Server
dataset_cache_dir_name

The name of the dataset cache directory, it is not recommended to modify
features_cache_dir_name

The name of the features cache directory, it is not recommended to modify
logging_level

Level control of Qlib-Server log
logging_config

Log configuration, it is not recommended to modify

Build from Dockerfile¶

Build Qlib-Server with Dockerfile according to the following processes:

Install docker, please refer to Docker Installation.

Start using Qlib-Server by running:

git clone https://github.com/microsoft/qlib-server
cd qlib-server
sudo docker build -f docker_support/Dockerfile -t qlib-server \
    --build-arg QLIB_DATA=/data/stock_data/qlib_data \
        QUEUE_HOST=rabbitmq_server \
        REDIS_HOST=redis_server \
        QUEUE_USER=rabbitmq_user \
        QUEUE_PASS=rebbitmq_password \
        FLASK_SERVER_HOST=127.0.0.1 \
        QLIB_CODE=/code
sudo docker run qlib-server

Using `Qlib` in `Online` Mode¶

Introduction¶

In the Qlib Document, the Offline mode has been introduced. In addition to offline mode, users can use Qlib in Online mode.

The Online mode is designed to solve the following problems:

Manage the data in a centralized way. Users don’t have to manage data of different versions.
Reduce the amount of cache to be generated.
Make the data can be accessed in a remote way.

In Online mode, the data provided for Qlib will be managed in a centralized manner by Qlib-Server.

Using `Qlib` in `Online` Mode¶

Use Qlib in online mode according to the following steps:

Open NFS Features in Qlib Client
Initialize Qlib in online Mode

Opening `NFS` Features in `Qlib` Client¶

If running on Linux, users need tp install nfs-common on the client, execute:
If running on Windows, do as follows.
- Open Programs and Features.
- Click Turn Windows features on or off.
- Scroll down and check the option Services for NFS, then click OK
Reference address: https://graspingtech.com/mount-nfs-share-windows-10/

Initializing `Qlib` in `online` Mode¶

If users want to use Qlib in online mode, they can choose either of the following two methods to initialize Qlib:

Initialize Qlib with configuration file
Initialize Qlib with arguments

Configuration File¶

The content of configuration file is as follows.

calendar_provider:
    class: LocalCalendarProvider
    kwargs:
        remote: True
feature_provider:
    class: LocalFeatureProvider
    kwargs:
        remote: True
expression_provider: LocalExpressionProvider
instrument_provider: ClientInstrumentProvider
dataset_provider: ClientDatasetProvider
provider: ClientProvider
expression_cache: null
dataset_cache: null
calendar_cache: null

provider_uri: 127.0.0.1:/
mount_path: /data/stock_data/qlib_data
auto_mount: True
flask_server: 127.0.0.1
flask_port: 9710

provider_uri

nfs-server path; the format is host: data_dir, for example: 127.0.0.1:/. If using Qlib in Local mode, it can be a local data directory.
mount_path

local data directory, provider_uri will be mounted to this directory
auto_mount:

whether to automatically mount provider_uri to mount_path during qlib init; User can also mount it manually as follows.

Note

Automount requires sudo permission
flask_server

data service host/ip
flask_port

data service port

Initialize Qlib with parameters configuration file as follows.

import qlib
qlib.init_from_yaml_conf("qlib_clinet_config.yaml")
from qlib.data import D
ins = D.list_instruments(D.instrumetns("all"), as_list=True)

Note

If running Qlib on Windows, users should write correct mount_path.

In windows, mount path must be not exist path and root path,
- correct format path eg: H, i…
- error format path eg: C, C:/user/name, qlib_data…

The configuration file can be:

...
...
provider_uri: 127.0.0.1:/
mount_path: H
auto_mount: True
flask_server: 127.0.0.1
flask_port: 9710

Arguments¶

Initialize Qlib with arguments as follows.

import qlib

# qlib client config

ONLINE_CONFIG = {
    # data provider config
    "calendar_provider": {"class": "LocalCalendarProvider", "kwargs": {"remote": True}},
    "instrument_provider": "ClientInstrumentProvider",
    "feature_provider": {"class": "LocalFeatureProvider", "kwargs": {"remote": True}},
    "expression_provider": "LocalExpressionProvider",
    "dataset_provider": "ClientDatasetProvider",
    "provider": "ClientProvider",
    # config it in user's own code
    "provider_uri": "127.0.0.1:/",
    # cache
    # Using parameter 'remote' to announce the client is using server_cache, and the writing access will be disabled.
    "expression_cache": None,
    "dataset_cache": None,
    "calendar_cache": None,
    "mount_path": "/data/stock_data/qlib_data",
    "auto_mount": True,  # The nfs is already mounted on our server[auto_mount: False].
    "flask_server": "127.0.0.1",
    "flask_port": 9710,
    "region": "cn",
}

qlib.init(**client_config)
ins = D.list_instruments(D.instrumetns("all"), as_list=True)

Note

If running Qlib on Windows, users should write correct mount_path.

The arguments can be:

ONLINE_CONFIG = {
    ...
    ...
    "mount_path": "H",
    "auto_mount": True,
    "flask_server": "127.0.0.1",
    "flask_port": 9710,
    "region": "cn",
}

Changelog¶

Here you can see the full list of changes between each QLibServer release.

Version 0.1.0¶

This is the initial release of QLibServer.

Version 0.1.1¶

Fix multi-process log bug, using logging.handlers.QueueHandler
Modify to create a rabbitmq connection and channel for each process, before sharing a one that causes basic_ack to be unsuccessful
Add max_process to config
The log is set through the configuration file, refer to config.yaml.example

Version 0.1.2¶

Format PEP8

Qlib-Server Documentation¶

Document Structure¶

Qlib-Server: Quant Library Data Server¶

Introduction¶

Framework¶

Listening to incoming request from client¶

Responding clients with data¶

Accepting tasks from RabbitMQ and processing data¶

Qlib-Server Deployment¶

Introduction¶

One-click Deployment¶

One-click Deployment with docker-compose¶

One-click Deployment in Azure¶

Step-by-step Deployment¶

Build RabbitMQ¶

Build Redis¶

Build NFS¶

Build Qlib-Server¶

Build with Source Code¶

Build from Dockerfile¶

Using Qlib in Online Mode¶

Introduction¶

Using Qlib in Online Mode¶

Opening NFS Features in Qlib Client¶

Initializing Qlib in online Mode¶

Configuration File¶

Arguments¶

Changelog¶

Version 0.1.0¶

Version 0.1.1¶

Version 0.1.2¶

`Qlib-Server` Documentation¶

`Qlib-Server`: Quant Library Data Server¶

`Qlib-Server` Deployment¶

One-click Deployment with `docker-compose`¶

One-click Deployment in `Azure`¶

Build `RabbitMQ`¶

Build `Redis`¶

Build `NFS`¶

Build `Qlib-Server`¶

Using `Qlib` in `Online` Mode¶

Using `Qlib` in `Online` Mode¶

Opening `NFS` Features in `Qlib` Client¶

Initializing `Qlib` in `online` Mode¶