Barcelona, Spain
October 25-28, 2016

Event Details

Chasing 1000 Nodes Scale

Performance working group is working on identifying best OpenStack performance verification since Mitaka, and this session is about one of our researches done during Newton release cycle. This research is targeting on how evaluating OpenStack workability and performance at large scale. In particular, we will discuss in this presentation 1000 nodes cloud emulation testing on top of two different testbeds. Scripts that have been used to perform those tests are freely available so that anyone can evaluate the performance of an OpenStack deployment either on different hardware as well as software parameters.

What can I expect to learn?

* Scalability limits of vanilla OpenStack with increasing number of compute nodes.

* Methodology / tools to benchmark OpenStack deployment.

* Default settings vs tuned settings (if required).

Tuesday, October 25, 12:15pm-12:55pm
Level: Intermediate
Matthieu Simonin is permanent research engineer at INRIA. His primary task is to help research teams to build and develop various software and experimentation on large infrastructures. His topics of interest include :  distributed systems, cloud computing IT automation Data visualisation FULL PROFILE
Dina has been working with OpenStack as a cloud platform for more than four years, both taking part in private cloud tuning and upstream contributing. Her experience includes close interaction with all OpenStack projects and their improvement to satisfy the needs of cloud users. Dina leads upstream Performance team initiative and Mirantis scale/performance testing initiatives. FULL PROFILE
Principal Engineer, Mirantis
Aleksandr joined the OpenStack community during the Folsom release. His main activity is helping companies adopt Openstack for development and production use cases. Aleksandr is guru of scale testing, he likes to move the limits forward by tuning and boosting the performance. FULL PROFILE