Cloud Scheduler for Next Generation DataCenters

Scheduler for next generation datacenters

Project logistics

Preferred past experience

Project Overview

A core part of the Massachusetts Open Cloud is a Hardware as a Service project to create a new fundamental layer in cloud datacenters. In a HaaS-enabled datacenter, researchers and developers can independently stand up services on bare hardware, with guaranteed isolation between them. They can scale their deployments by allocating and freeing hardware nodes, without having to change any physical setup of the datacenter. HaaS allows bare metal machines to be used similarly to virtual machines in the cloud. We want to explore two models for scheduling projects using HaaS. First, a simple, traditional scheduler (e.g., SLURM). Second, one based on the Mesos scheduler for supporting multiple frameworks. Imagine a hybrid cloud data center with one framework consisting of thousands of servers hosting XBOX live applications and the remaining frameworks hosting various applications such as batch-style scientific computations and interactive Office 365 sessions. Based on load and priorities among different frameworks, the servers should be able to support these very different applications moving resources between them based on demand as swiftly as possible, and the scheduler must make those decisions dynamically. The work will interact heavily with both HaaS and the BMI (Bare Metal Imaging) project The Image Management Service (IMS) of BMI will play a prominent role.

Some Technologies you will learn/use: