https://github.com/emc-advanced-dev/layerx

The Framework for Cross-Cluster Scheduling
https://github.com/emc-advanced-dev/layerx

Last synced: 10 days ago
JSON representation

The Framework for Cross-Cluster Scheduling

• Host: GitHub
• URL: https://github.com/emc-advanced-dev/layerx
• Owner: emc-advanced-dev
• License: apache-2.0
• Created: 2016-09-09T18:45:44.000Z (about 8 years ago)
• Default Branch: master
• Last Pushed: 2017-02-02T19:55:08.000Z (almost 8 years ago)
• Last Synced: 2024-08-02T02:15:08.272Z (3 months ago)
• Language: Go
• Homepage:
• Size: 15.8 MB
• Stars: 72
• Watchers: 12
• Forks: 8
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
# 

**Layer X** a single centralized service that manages containers across multiple clusters, multiple clouds, and multiple cluster managers. it joins together pools cluster managers such as [*Kubernetes*](http://kubernetes.io/), [*Docker Swarm*](https://docs.docker.com/swarm/), and [*Mesos*](http://mesos.apache.org/) into a single shared group of resources, where any container can be managed, scaled, migrated across a unified resource pool. 

**Layer X** exposes a set of cluster-management APIs including scheduling & resource management to so-called “brains” which enables advanced features such as live migration & continuous placement of containers, easy flex up & flex down of cluster, and dynamic resizing of resource constraints.

To learn more about the motivation behind project Layer X, watch our [slides](http://www.slideshare.net/IditLevine/layer-x-oss-65996486) and watch a [demo](https://youtu.be/YEqYIm-jbB4).


To stay up-to-date on Layer X, join us on our [slack channel](http://project-unik.io).

![Layer-X Arch](http://i.imgur.com/O5TxJLF.png "Architecture")

In the **Layer-X** world:

* **Resource Providers** offer nodes with useable resources (e.g. a [Mesos Slave](https://open.mesosphere.com/reference/mesos-slave/), a [Kubelet](http://kubernetes.io/docs/admin/kubelet/), or a [Swarm Node](https://docs.docker.com/engine/swarm/swarm-tutorial/add-nodes/)) to be consumed by

* **Task Providers**, which create services and perform work by launching Docker containers (e.g. [Marathon](https://mesosphere.github.io/marathon/), [Deis](http://deis.io/))

* **Brains** interface with the **Layer-X** HTTP API to manage the global state of the cluster, including scheduling *tasks* on *nodes*.

**Layer-X** abstracts specific implementations of **task providers** and **resource-providers** with **Task Provider Interfaces** (TPIs) and **Resource Provider Interfaces** (RPIs). Currently, TPI and RPI have only been completed for Mesos, but Kubernetes and Docker Swarm will be available soon™.

**Layer-X** implements a simple interface for adding TPIs and RPIs to encourage easy community contributions for new resource providers (Nomad, Hadoop, etc.) and task providers (Dokku). Client APIs for TPIs are visible [here](./layerx-core/layerx_tpi_client/layerx_tpi.go) and for RPIs [here](./layerx-core/layerx_rpi_client/layerx_rpi.go).

**Brains** are also designed to be pluggable, 3rd party extensions for **Layer-X**. The Brain API is visible through the Brain Client package [here](./layerx-core/layerx_brain_client/layerx_brain_client.go)

Manually install with the [installation documentation](docs/install.md), or bring up a local cluster in minutes with [vagrant](vagrant/).

---

A rough diagram of Layer-X architecture:

![Layer-X Arch Diagram](http://i.imgur.com/GcYh5ug.png "Architecture")