In both NFV and SDN scenarios, most of the VM\Container cannot work is attributable to the cause of network disconnection, packet loss, latency, and so on.
These problems are really everywhere, and the consequences are less predictable. SDN might has some sloutions, but SDN tools may not be perfect, and SDN only slove network not for VM\Container, also some of the network is not SDN.
Further said, according to a large heterogeneous cloud environment, the most simple goal is to restore all of the VM\Container rather than just network fault recovery.
So we will build a auto-healing closed-loop framework for VM\Container: find fault, analysis, decision-making, workflow recovery actions.
And the framework of our design we will integrate skydive for network data collection and fault detection.
This will be a huge challenging and exciting practice, we'll always go on.
In this Presentation, we will take you to the world of the following:
1 Build a auto-healing closed-loop framework for VM\Container: find fault, analysis, decision-making, workflow recovery actions.
2 In the framework of our design we will integrate skydive for network data collection and fault detection.
3 Visual all ports and links state about VM\Container and BareMeter, let cloud maintenance personnel for network condition be clear at a glance.
4 Also we will demonstrate a demo with network fault expert database to show how to auto-healing network fault; If network fault cannot be auto-healing, we will how to auto-healing VM\Container; If VM\Container also can't auto-healing then how to deal with.