backend.ai 1.0.1

Lablup Backend.AI Meta-package

Backend.AI is a streamlined backend service framework hosting heterogeneous programming languages and popular AI frameworks. It manages the underlying computing resources for multi-tenant computation sessions where such sessions are spawned and executed instantly on demand.

In the names of sub-projects, we use a private code-name “Sorna” which we take from a famous science fiction “Jurassic Park” – meaning that we do all the dirty jobs in the behind. In the novel, Isla Nublar is the “front-end” island where tourists see the dinosaurs and Isla Sorna is the “back-end” island where a secret dinosaurs production facility is located.

All sub-projects are licensed under LGPLv3+.

Server-side Components

Manager with API Gateway

It routes external API requests from front-end services to individual agents. It also monitors and scales the cluster of multiple agents (a few tens to hundreds).

• Package namespace: sorna.gateway, sorna.manager

• https://github.com/lablup/sorna-manager

Agent

It manages individual server instances and launches/destroys Docker containers where REPL daemons (kernels) run. Each agent on a new EC2 instance self-registers itself to the instance registry via heartbeats.

• Package namespace: sorna.agent

• https://github.com/lablup/sorna-agent

REPL

A set of small ZMQ-based REPL daemons in various programming languages and configurations. It also includes a sandbox implemented using ptrace-based sytem call filtering written in Go.

• https://github.com/lablup/sorna-repl

• Each daemon is a separate program, usually named “run.{lang-specific-extension}”.

Sorna Common

A collection of utility modules commonly shared throughout Backend.AI projects.

• Package namespaces: sorna.common

• https://github.com/lablup/sorna-common

Client-side Components

Client Libraries

A client library to access the Sorna API servers with ease.

• Python

  • pip install backend.ai-client

  • https://github.com/lablup/backend.ai-client-py

• Javascript (under preparation)

  • npm install backend.ai-client

  • https://github.com/lablup/backend.ai-client-js

• PHP (under preparation)

  • composer require lablup/backend.ai-client

  • https://github.com/lablup/backend.ai-client-php

Sorna Media

The front-end support libraries to handle multi-media outputs (e.g., SVG plots, animated vector graphics)

• The Python package (lablup) is installed inside kernel containers.

• To interpret and display media generated by the Python package, you need to load the Javascript part in the front-end.

• https://github.com/lablup/sorna-media

Integrations with IDEs and Editors

Sorna Jupyter Kernel

Jupyter kernel integration of the Sorna Cloud API.

• Package namespaces: sorna.integration

• https://github.com/lablup/sorna-jupyter-kernel

Visual Studio Code Extension

Extension for Visual Studio Code to run your code on the Lablup.AI clouds or your own Backend.AI servers.

• Search ‘live code runner’ to VSCode extension search.

• https://github.com/lablup/vscode-live-code-runner

Atom Editor plugin

Atom Editor Plugin that allows running your code on the Lablup.AI clouds or your own Backend.AI servers.

• Search ‘live code runner’ to Atom plugin search.

• https://github.com/lablup/atom-live-code-runner

Installation

The Sorna project uses latest features in Python 3.6+ and Docker CE 17.05+.

To install the manager with API gateway, run:

pip install backend.ai[manager]

For each computing servers, install the agent using:

pip install backend.ai[agent]

NOTE: More details about configuration will be released soon.

Development

git flow

The sorna repositories use git flow to streamline branching during development and deployment. We use the default configuration (master -> preparation for release, develop -> main development, feature/ -> features, etc.) as-is.