Large scale and batch computing on Google Compute Engine (Google Cloud Next '17)

Опубликовано 10 марта 2017, 4:43

In this video, you'll learn how to think about and architect batch processing systems on Google Compute Engine (GCE). Michael Basilyan and Eric Brewer cover reference architectures applicable to business, engineering, and scientific applications as well as how to efficiently leverage GCE instances, networks, and disks. They also discuss best practices for scaling your workloads up to tens of thousands of cores. You'll learn cost-saving strategies, such as the use of Preemptible VM instances, so you can put more resources towards innovating for your customers. Michael and Eric are joined by Dr. Burt Holzman, coordinator for Scientific Computing Facilities, Fermi National Laboratory. Dr. Holzman discusses how Fermi was able to scale to 160,000 cores on GCE for High Energy Physics research.

Missed the conference? Watch all the talks here: goo.gl/c1Vs3h
Watch more talks about Infrastructure & Operations here: goo.gl/k2LOYG