Event Details

As we presented at the last OpenStack Summit in Austin, we shared how to build real-time cloud using OpenStack, proposing a reference architecture. And now it seems that more users and operators need to run low-latency or real-time applications (such as audio/video streaming, highly interactive systems, IoT, HPC, etc.) to meet their requirements in clouds.

This time we we share how to analyze and manage performance issues that users or operators can encounter, showing a framework on performance monitoring.  Since we are working on Real-time KVM as one of the projects of KVM enhancements for NFV at OPNFV, we share our know-how and experiences to root-cause performance or real-time issues on OpenStack. We use more common cases to make the presentation comprehensive and generic. Typical NFV workloads push the limits in all of the system resource areas: CPU, I/O, and memory, and our our know-how and experiences should be applicable to common cases.