High service availability is an integral part in mobile core network. As each core node hosts few hundred thousand subscriber sessions, a sub-second level fast recovery is necessary not only to ensure seamless connectivity to the customers, but also to avoid the consequent ‘attach’ storm where all disconnected mobile devices try to reconnect to the mobile network simultaneously, creating control traffic beyond what a node can handle. In NFV-based networks, recovery actions will be performed by a node manager which cannot directly detect virtualized network infrastructure level failures. Therefore, OpenStack needs to detect and notify the corresponding node manager of relevant resource failure. An additional requirement is the filtering of the failures, as the definition of failure and the operational policy may differ among Operators.
We’ll introduce a solution using Congress and extended features to meet these telco requirements, as well as PoC developed with OPNFV Doctor Project.
The Telecom/NFV track is brought to you by OpenStack and OPNFV.
- Details use cases and the consequent requirements for failure management in a mobile cellular network, from NFV-based network design, development and deployment experience on OpenStack
- How Congress can play a vital role in meeting such failure management requirements
- On-going progress and success story of collaboration among operators, telco-vendors and open source developers in OpenStack and OPNFV