Voting Just Opened for OpenStack Australia
The 2017 Openstack Summit just opened up voting for presentations to be given Nov. 6-8, Sydney, Australia. Mellanox has a long history of supporting OpenStack with technology and product solutions. Mellanox has submitted a number of technical papers and would like to urge readers to vote, vote, vote! The OpenStack Foundation receives more than 1,500 submissions and of these, they only select 25-35 percent for participation so your votes count.
In this blog, we will cover the topic and content of three of our submissions. There is only one short week to vote so please review the content and vote for your favorites. Voting begins July 25, 2017 and ends August 1, 2017.
First paper topic: Achieving The Best of Both Worlds: Running Highly Efficient Containers Inside High Performance VMs
Overview: Running containers within VMs (a.k.a. nested container) is a topic for debate. There are pros and cons associated with such a deployment model. Nested containers alleviate two major challenges; security/isolation and lack of popular orchestration tools. However, they solve these problems by degrading network performance drastically.
Traditionally, open vswitches running in a hypervisor have a poor network I/O performance. With nested containers, this performance degradation problem gets compounded even further. In addition to the software switch in hypervisor handling packets to various VMs, there is another software switch that is running in each VM that dispatches packets between the containers within VMs.
In this presentation, we will present ways to accelerate packet switching at the hypervisor and VM level to improve performance of nested containers by an order of magnitude.
Second paper topic: Openstack Storage – Born to be fast!
Overview: OpenStack automates storage provisioning, but what about storage performance? As servers and drives get faster, how can cutting edge storage solutions remain flexible and deliver high throughput, low latency and high IO operations per second to demanding cloud applications?
In this presentation, Mellanox will explore three ways to accelerate OpenStack storage. All three utilize a standard, advanced network transport called Remote Direct Memory Access (RDMA), which increases storage performance and lowers CPU utilization. Highlights include:
Third paper topic: Brain in the cloud: Machine Learning On OpenStack Done Right!
Overview: Machine Learning is, no doubt, the hottest trend in information technology these days. Deep Neural Network (DNN), a subfield of Machine Learning with structure and mode of operation loosely inspired by the brain, allows technologists to solve complex problems such as image recognition that has been very difficult or even impossible to solve using standard programming paradigms. Mellanox therefore believes that DNN will grow in popularity to become a standard tool for many applications.
DNN concepts are not new. However, and until recently, applying them in practice could not be realized due to their high computational demands. With the recent development in parallel computing, especially around GPU acceleration and high speed and efficient networking, DNN has become a reality in modern data centers.
In this presentation, Mellanox will describe the system requirements to effectively run a machine learning cluster with popular frameworks such as TensorFlow. This will include a discussion regarding how such a system can be deployed in an OpenStack-based cloud without compromises, enjoying high-performance DNN programming paradigm as well as the benefits of cloud and software-defined data centers. Finally, Mellanox will highlight a case-study from Monash University.
Mellanox has contributed to the open source community in all three areas and supports these solutions.
Thanks for voting and see you in Sydney!