System and method for state replication of virtual network function instances

Abstract

A method to be implemented by an interface element (IE, 510) between a VNF pool (508) and a control entity (520) is disclosed. The VNF pool (508) includes a plurality of VNF instances (502), including at least first and second instances. A control entity (520) is configured to control the VNF instances (502) of the pool (508). The method enables the IE (510) to help replicate the state of the first VNF ​​instance on the second VNF instance. The method includes the steps of: obtaining a control message provided from a control entity (520) to the first VNF ​​instance (502), providing the obtained control message to at least the first VNF ​​instance and the second VNF instance (502), and when the interface The element (510) provides control to the controlling entity (520) when it has an explicit or implicit acknowledgment of both the control message provided to the first VNF ​​instance (502) and the control message provided to the second VNF instance (502) Acknowledgment of the message.

Description

technical field [0001] The present disclosure relates generally to the field of virtual network functions (VNFs). In particular, although not necessarily, the present disclosure relates to methods, systems and computer program products for replicating the state of one VNF instance of a pool of VNFs on at least one other VNF instance of the pool. Background technique [0002] In conventional telecommunications networks, functions such as firewalls, load balancers, and radio access network (WAN) optimizers are deployed on dedicated specialized hardware servers in the networks of network operators and data centers. These functions are important because they serve as the building blocks of all web services. A recent trend is to virtualize these functions, ie to decouple them from their dedicated hardware by implementing the function as a virtual network function instance (VNF instance) eg as software running on a general-purpose server. [0003] One challenge of deploying netw...

AI Technical Summary

Problems solved by technology

The number n of rules in an OpenFlow switch can easily be on the order of 10K flow entries, which represents state for more than a few hundred MB of memory, which results in significant failover times

[0014] Although the above discussion has mainly focused on the failover process, similar situations arise in situations where the state of one VNF instance of a VNF pool needs to be replicated to at least one other VNF instance of the pool in the absence of failure of the primary VNF. question

Images