Thanks to all who joined us for the Nokia NFV VNF Report Webinar: Nokia Cloud Native for Fully Digitized Businesses, where Nokia discussed cloud native design, creating a Telco cloud and leveraging the full potential of cloud in Telco. After the webinar, we took questions from the audience but unfortunately ran out of time before we could answer everybody’s questions. You can read the full Q&A below.
Are the initial benefits we get from virtualization and the early stages of NFV enough, like HW independence and better scalability? Is cloud-native really a necessary step for the VNFs?
The first step on the journey to Telco Cloud and NFV was the virtualization of the network elements. With virtualization, the industry was able to achieve some of the benefits expected from the cloud architecture like HW independence, better scalability and an increased degree of flexibility, however this is clearly not enough. To be able to leverage the full potential of cloud in telco we need to continue the evolution towards implementing a cloud-native architecture. That passes through several steps, as outlined in the presentation and the whitepaper shared with the audience.
Ultimately, the adoption of a cloud-native core network will enable the flexibility, responsiveness and adaptability needed to deliver the high performance, ultra-reliability and low latency demanded by 5G and IoT.
Scenario: An operator has implemented a unified data repository for subscriber data management in bare metal- would it be possible and would it make sense to implement Nokia Shared Data Layer to introduce session data?
Yes, it would make great sense to do that. There is no absolute need to implement both subscriber and session data on Shared Data Layer at once. An operator may decide to leave the subscriber data on the existing systems and deploy Shared Data Layer just for Session data. In this scenario the operator will already have the benefits of Shared Data Layer and can decide to migrate subscriber data to the new platform in another opportunity.
What are the particular challenges of implementing DevOps on the Telco environment?
First on the SW design as most of the Telco VNFs are still deployed on a monolithic architecture which make the continuous delivery of SW difficult. The further disaggregation of the VNFs with the potential adoption of micro-services is a necessary step to simplify SW delivery and update procedures and thus facilitate the adoption of DevOps.
Second, and most importantly, there is also a need to define how DevOps will work on a scenario where R&D sits on the vendor organization and operations owned by the CSP. The industry will need to update processes and further evolve the relationship between vendors and operators to enable the level of collaboration needed to make it feasible to implement a continuous SW delivery and feedback cycle.
Looking at IoT and the massive number of devices, would it make sense to introduce cloud-native for IoT as an overlay solution?
Definitely. Introducing cloud-native core for the new applications like IoT makes absolute sense. Some characteristics of the cloud-native core will contribute significantly to optimize the utilization of network resources on some of the IoT use cases.
For example, on a very basic IoT scenario like a widespread implementation of smart meters, the ability to check point data session information on the Shared Data Layer allows the network to avoid dedicating packet core resources to the idle sessions, allowing the infrastructure to support much more devices than it would originally be able to.
Where is the industry in terms of standardization of the SDL? Are there interfaces currently defined?,
Motivated by the very positive feedback from operators and analysts, when launching Shared Data Layer beginning of 2016, Nokia proposed Shared Data Layer as a 3GPP 5G core study item with the outcome that it will be part of 3GPP release 15 standardization under the term of “Data Storage Function, DSF”.
Shared Data Layer uses standard interfaces as for example LDAP, CQL, SOAP.