Vancouver, BC
May 21-24, 2018

Event Details


Closing the Loop: VNF end-to-end Failure Detection and Auto Healing

Openstack Vitrage provides a framework for analyzing and expanding OpenStack alarms & events. In this session we'll demonstrate how to close the loop between cloud monitoring and maintaining VNF live and confirmed states, using a generic VNF Manager. This closed loop allows operators to expand their limits and perform automatic actions such auto-healing or auto-scaling. In a live demo we'll show how the generic VNFM uses its auto operations capability to sample the VNFs health, and report its state to Vitrage.  Vitrage is responsible for correlating the HW failures with the VNF resource health, and through a process of deduced alarms, notifying the VNF Manager on VNF failure. Once we achieve a closed loop, we'll show the centralized dashboard in the Vitrage OpenStack UI, as well as VNF healing performed by the VNF Manager using the Mistral OpenStack workflow service to act on Vitrage insights.


What can I expect to learn?

Audience will learn about VNF fault management using OpenStack project. Demo and presentation will show end to end from failure detection to VNF auto healing.

Tuesday, May 22, 11:50am-12:30pm
Level: Intermediate
Nokia Cloudband
Ohad is a product manager in CloudBand, Nokia. In his role, Ohad is leading Analytics and monitoring for the CloudBand NFVI product and he is also responsible for open source activity in CloudBand. Ohad is driving Vitrage, an official OpenStack project, initiated by CloudBand, for root cause analysis, deduced alarms and states. FULL PROFILE
nokia
Product manager for Nokia's VNF Manager offering (CBAM- Cloudband application manager). FULL PROFILE
Software Developer
Software Developer in CloudBand, Nokia FULL PROFILE