With more than 80 production deployments across more than 45 communication service providers (CSPs), you can safely say that we have gained significant real-world experience in deploying and supporting our customers in rolling out NFV platforms, creating new services, and operating these services. Today we announce the new version of our VMware vCloud NFV platform, which represents the culmination of this experience and our ongoing commitment to supporting our service provider customers in reaching their business goals.
Our customers, and the industry overall, have been telling us they are looking for three specific capabilities:
- They want service agility and quicker time to revenues. The promise of NFV is firmly rooted in the ability of CSPs to deliver services faster with great agility and significantly less costs.
- They demand reliability and availability. The services that have been or will be virtualized must operate with the same or better quality as existing services today. This is especially true for services considered to be critical infrastructure or highly regulated.
- They want choice. They want platforms that plug-into some of their existing support systems and they want to be able to choose freely between Virtual Network Function (VNF) vendors as well as MANO suppliers. You may even say that no two CSPs are exactly alike.
So what do we have in store for VMware vCloud NFV 2.0? Let’s take a look
Network Modernization With VMware vCloud NFV 2.0
The foundation of any NFV platform is a robust, carrier-grade infrastructure. VMware vCloud NFV 2.0 excels in this area, offering network functions virtualization based on:
- VMware vSphere 6.5 – the defacto industry standard in compute virtualization.
- VMware NSX for vSphere 6.3 – the industry’s most widely deployed, production-proven network virtualization platform
- VMware vSAN 6.5 – arguably one of the most efficient ways to virtualize storage in any NFVI.
All three components provide an quantum leap forward towards increasing performance, high availability, security and scale of the NFVI.
With this release, CSPs gain fine-grain control of resources assigned to VNFs, such that when performance is needed, the VNF benefits from dedicated resources. We also include native Intel drivers in our hypervisor, increasing packet processing efficiency. VNFs can now benefit from additional High Availability features that are built into the platform directly and can be used by any VNF. Our platform can detect physical server health degradation (from a selected number of market-leading server vendors such as Dell, Cisco and HP) and vacate VNFs from the damaged server. When VNF components must be restarted in certain order, a new feature called Orchestrated HA allows the operator to dictate the order in which components are restarted. For those workloads considered security sensitive, we now offer VNF component-level encryption and secure boot.
With these, and many of the additional features in vCloud NFV 2.0, we increase the robustness and performance of the NFVI, but as any operator knows, none of these capabilities are sufficient if the platform is difficult or obtuse to manage.
This is the second pillar on which vCloud NFV 2.0 stands. In addition to the powerful VMware vRealize Log Insight and VMware vRealize Operations Manager capabilities, we have added VMware vRealize Network Insight. With these three components, we now can monitor, manage and operate every layer in the virtualization stack: from the components of the VNF all the way to the interaction between the NFVI layers. The three components help eliminate the mystery between the layers and provide a holistic view into the entire environment. VMware vRealize Operations Manager can also be extended with plugs ins to monitor key performance indicators directly from the VNFs themselves.
Imagine that your network operations center (NOC) can monitor your complete NFV infrastructure from a single pane of glass – yeh, we got that. The capabilities of our integrated and extensible operations management components enable CSPs to reduce operational cost through fast Mean Time to Understand and Repair (MTTU/MTTR), contextual troubleshooting and visual root-cause analysis. In fact, when investigating an issue, the operator could use Log Insight’s search tool in much the same way that a google search is done, and then continue with the analysis using network insight’s powerful visualization. Even communication between VNF components and their interaction with external physical elements is supported.
A robust and well managed platform is the foundation of the virtualized CSP. But these two attributes are still insufficient to quickly roll out new services from a best-in-class selection of suppliers,and to expand on these services as they grow.
That’s why VMware has been working furiously with a myriad of VNF vendors to test and verify functional interoperability between their VNFs and vCloud NFV components. We’re proud to share that in the last 8 months we have nearly tripled the number of certified solutions to a total of 34 in all network areas, with even more on the way. Partners are willing to work with VMware, because they see the significant “skin in the game” we have through our investment in certifying their solutions. VMware’s own engineers and lab are responsible for the testing; the participating vendors do not need to invest more effort. The VMware Ready for NFV program not only provides CSPs with choice in each of the functions they’d like to virtualize, it increases confidence in their ability to compile and deploy a multi-vendor solution. Partners who have completed the certification program, called VMware Ready for NFV, can be found here in our Solutions Exchange.
The next step in a typical CSP service preparation (after all the legal and financial aspects have been addressed) is on-boarding the service components into a Virtualized Infrastructure Manager (VIM). With vCloud NFV 2.0, we upgrade both VIMs (remember, our customers want choice). VMware vCloud Director 8.20 is even more tightly integrated with our NFVI stack and we have updated hte platform to VMware Integrated OpenStack 3.1, which is based on OpenStack Mitaka.
Both VIMs can easily consume VNFs using vApp (for vCloud Director) and Heat templates for VIO. Importing the templates into the VIM can be automated, and with the new releases, significant enhancements have been made to the vApp template such that affinity, NUMA, performance and policies could be configured. To make any service truly scalable as well as quality-assured, customers (tenants of the platform) can be guaranteed certain amount of resources that include not only compute and storage, but also bandwidth. In fact, the level of secure multi-tenancy that’s introduced with vCloud NFV 2.0, can enable a whole new set of services. Communication service providers can now use a fine-grain role based access control (RBAC) to assign maintenance responsibilities of the VNF to its supplier, and local administrative rights to the customer’s network admin. Imagine the possibilities.
Another prevalent use case, for more general purpose multi-tenancy, is the ability to slice network services and resources between VNFs and locations. Especially this week, as 5G, NB-IoT and network slicing is taking over the headlines, the ability to support such use cases quickly and with less risk is essential. In our vCloud NFV 2.0 demo at Mobile World Congress, we show how such scenarios can exist.
Are we excited? You bet. Is the list above all inclusive, no way.
With vCloud NFV 2.0 we empower our customers to accelerate their services, simplify their operations and provide unparalleled choice. As more and more NFV platforms are deployed, network functions are virtualized, and CSPs are provided with the tools and knowledge to embrace this transformation, we will continue to collaborate with our customers to help them realize their business goals and benefit from our experience in an industry we helped create.