Blog written by Stefan Vallin, Principal Engineer, Cisco Systems
Consider the three key steps related to provisioning and maintaining a customer network service. First, configuration of network devices according to the service intent, followed by activation tests. Finally, ongoing service level assurance monitoring to verify that the service delivered is what was committed.
Up until now, these activities have lived completely isolated from each other, in separate silos. Not surprising, it’s just human nature to break down complex systems and processes into separate entities since it makes things easier to manage, but the time is now to collapse these silos.
It’s been the OSS’s role to align these three activities. A task doomed to fail as traditional stovepipe architectures haven’t been designed to share knowledge of the overall network service to be provisioned, causing considerable frustration in the industry.
Cisco, together with Netrounds, Intel, Orange and Telco Systems are now innovating this space and bringing the results through a “Proof of Concept” carried out within the ETSI NFV Industry Standardisation Group. The key enabler to make this possible is network programmability, combined with active software-based virtual probes capable of assuring the end-user experience.
The PoC aims to show orchestration of a network service through Cisco Network Service Orchestrator (NSO) including each of the three aspects described above by using a tight service model coupling based on YANG. The PoC also aims to show that service KPIs can be bound to the orchestrated service and monitored by active virtual probes, so-called vProbes. In the PoC, these vProbes are provided by Netrounds, and are launched on NFV infrastructure, both in the datacenter environment, as well as on the demarcation locations at so called distributed NFV locations.
Cisco NSO is not only configuring the VNFs representing the service, but it is also performing sibling provisioning of virtual probes that both test and monitor the service actively. A YANG model is used as data-model language both for the NFVO MANO descriptors and for the OSS, covering both fulfillment and assurance.
Referring to network programmability, Cisco NSO utilizes a cloud-based API towards Netrounds’ Control Center, which acts as a unifying server towards all the vProbes included in the service chain. Referring to the service chain, this also includes a virtual content cache, as well as Cisco’s virtualised PE router.
The main use case is driven by Orange and aims to verify that using content caching closer to the end-user location, the user experience of HTTP-based streaming video will be improved.
Along with our virtual managed services solution that we discussed last week, we’ll be showcasing the PoC at Cisco’s booth at the Big Telecom Event in Chicago, June 9-10, 2015. Please come by and we will show you the demonstration. We will have Netrounds joining us at the booth to share their insights in the future of joint service orchestration and assurance using virtual probes.
Get rid of your frustration and start collapsing your OSS silos. We’re here to help you through the transformation!
Tweet us @CiscoSP360 if you have any questions or comments.