Manage the future now

Business white paper

Manage the

future now

Table of contents

1 Executive summary

1 Structure of this document

2 NFV business goals and challenges

4 Leveraging NFV for OSS transformation

7 HP vision for OSS transformation

14 Implementing NFV management in OSS

17 Conclusions

Network functions virtualization (NFV) is not a matter of if or

when anymore. It is indeed a matter of how.

Executive summary

Communications service providers (CSPs) are required to streamline their cost structures and operations while increasing portfolio competitiveness and drastically slashing time to market. Network functions virtualization promises all this and more.

As often indicated by operational teams, however, NFV will have a significant impact on CSP operations support systems (OSS) and business support systems (BSS). This will require at least revisiting, if not a major redesign, to keep pace with the automation, elasticity, capacity optimization, and scalability advantages allegedly brought by network functions virtualization. As a consequence, NFV adoption means OSS transformation, with all the challenges that such a journey will entail. Whether a radical change or a gradual approach is chosen to address the OSS component of the NFV deployment, service providers will have to look at designing a target management architecture—derived by current approaches, such as TM Forum Business Process Framework (eTOM), but extended to address NFV, based on solid pillars—to guarantee a futureproof journey and a smooth transition from the current situation, systems, platforms, and processes. The HP OSS vision is for service providers to break the OSS silos and migrate to a dynamic production environment, based on the concept of service operations factory (SOF). This is where the integration between fulfilment and assurance will create a virtuous circle to make service management more agile, effective, and complete.

At the same time, the service operations factory will drastically reduce time to market and improve end-to-end service visibility. Completed by data analytics functionality, such an approach also is designed to align OSS capability with the benefits promised by the NFV implementation in terms of cost-effectiveness and service management agility. It offers service providers a future-proof platform and a smooth transition to the hybrid virtual and nonvirtual world independently from the direction the NFV journey will take them.

Structure of this document

This paper presents the promises of network functions virtualization and the associated infrastructural, operational, and organizational challenges that come with it. It also will present the demands of NFV for OSS transformation, as well as the paradigm shift of resource management toward service management. In addition, it explores the impact of NFV on OSS fulfillment and assurance and shows how NFV can be embedded in operations support systems, without requiring a big-bang approach.

The HP vision for OSS transformation is presented by:

• Leveraging the service operations factory and OSS & NFV analytics to provide and experience new technologies for the OSS of the future

• Completing the NFV management and orchestration (NFV-MANO) capabilities provided by HP NFV Director

Network functions virtualization (NFV) is a core structural change in the way telecommunications infrastructure is deployed. This, in turn, will bring significant changes in the way that services are delivered by communications service providers (CSPs)—and hence have an important impact on operations support systems (OSS).

2

The following convention will be used throughout this document: Whenever a concept has a specific definition given by standard bodies, such as the European Telecommunications Standards Institute (ETSI), the term will be capitalized. If the general English meaning is intended, the word(s) will be in lower case. For example, “Network Service” refers to its ETSI definition, whereas “network service” refers to its general meaning in English.

NFV business goals and challenges

NFV is a core structural change in the way telecommunications infrastructure is deployed. This, in turn, will bring significant changes in the way that services are delivered by service providers.

Business goals enabled by NFV

NFV transformation promises to bring major business, technology, and operational benefits to increase carriers’ competitiveness.

What is expected by the adoption of NFV can be broken down into:

• Optimized operating expenses (OPEX): Move from the telco OPEX model to a lower cost IT

OPEX model, enabled by the reduction in unit cost of infrastructure that results from the use of industry standard platforms. It also reduces support cost for maintaining the infrastructure because of better sourcing availability,

• Lower capital expenses (CAPEX): Move from dedicated appliances to the virtualization

model, based on IT technology. This model is enabled by the standardization of service and the Virtual Network Function (VNF) lifecycle, the simplified and automated operations of uniform, homogenous physical platforms, and the decreased capital investment coming from increased capacity utilization.

• Accelerated time to market: Deploy new software and services quickly and easily (from

months to minutes), enabled by DevOps service creation, rapid prototyping and testing, and the ease of deploying remotely and automatically software to address service needs.

• Accelerated innovation via an open platform: Broaden access to partners that can innovate,

not just network equipment providers (NEPs) but also startups, software vendors, and others enabled by open interface adoption and true multivendor ecosystem.

• Delivered business agility: Rapidly scale up or scale down applications, modify quality of

service (QoS), and deliver new services faster.

• New opportunities: Service providers will be able to benefit from new possibilities to

increment their portfolios with services directly related to the virtualization, such as disaster recovery and flexible capacity.

• Improved business processes: From a management viewpoint, virtualization facilitates

decoupling between applications and the underline infrastructure, decreasing some of the challenges service providers face today, such as those related to inventory accuracy. This will provide CSPs with the unique opportunity to evolve in a more efficient mode of operations, anticipating the virtualization needs while leveraging the improvement of operational capabilities end to end.

Short-term, OPEX improvements are the critical point because they’re the only thing that could hope to be quantified right now.

NFV challenges

Even though NFV is promising unique benefits for service providers, it will impact future organizations and create essential challenges for operations and OSS. These challenges in turn might impact chances of a successful transformation. They can be classified in three categories: infrastructural, operational, and organizational.

Infrastructural challenges

Infrastructure challenges come from the introduction of new types of components in the network, originating from the IT world and based on industry standard platforms. However, these components need to provide telco-grade availability, as well as the needed performance and meet service-level agreements (SLAs). NFV also has introduced uncertainty as it is still not mature; this mandates a future-proof architecture to cope with it.

From a technology perspective, NFV is introducing a new abstraction resulting from the indirection between network functions and resources. It adds complexity for the existing OSS landscape and requires additional orchestration. It creates an elastic relationship between services and resources that makes SLAs and problem resolution much more difficult. Existing classic network functions will be replaced by VNFs, with their own dynamics and elasticity (such as scaling in or out), potentially instantiated on different data centers, and following various affinity and location rules. Standard network devices are becoming VNFs, which can be considered to be services. This necessitates a paradigm shift in the OSS: from resource-centric management to service-centric management.

Operational challenges

One of the main operational challenges introduced by NFV is how to maintain customer and services views that are correlated to the underlying infrastructure. This requires integration within existing OSS/BSS environment and end-to-end automation to enable NFV agility and faster service velocity. The other important operational challenge is how to ensure reduced operations costs while deploying NFV—with its associated complexity, changes in lifecycle processes, and the need for correlation and automation at all levels.

Organizational challenges

The last set of challenges facing service providers when deploying NFV is organizational. Not only will the change in processes have operational impact but it will require that IT and network operations organization cooperate better. The planning and building of the network will change from focusing on dedicated resources to capacity planning in the NFV infrastructure (NFVI), which will have significant impact on the planning and purchasing departments. Finally, the agility introduced by NFV, with its paradigm shift toward services, also might have direct impact of the operations organization, changing the way it does the job, and possibly requiring the creation of new job profiles that sit across the chief technology officer (CTO) and chief information officer (CIO) domains.

Need for NFV orchestration

From technology perspective, NFV is introducing a new layer for virtualization. It adds

complexity for the existing OSS landscape and requires additional orchestration to manage the new lifecycle of VNFs and Network Services (NS).

As shown in Figure 1, a number of existing classic network functions will be replaced by VNFs, with their own dynamics and elasticity.

BGCF MGCF IMS AS CSCF MSE MRF

Network

functions

IP

layer

Optical

transmission

Virtual

network

functions

IP

layer

Optical

transmission

MRF

4

This additional complexity needs to be managed but in a way that provides automation that delivers the agility needed for faster service enablement.

As identified by the European Telecommunications Standards Institute (ETSI) NFV Industry Specification Group (ISG), these challenges require new NFV orchestration functions, called the NFV Management and Orchestration (NFV-MANO) framework shown in Figure 2.

Figure 2. The NFV-MANO architectural framework

The NFV-MANO handles the lifecycle of Network Services and VNFs. It brings a new set of management capabilities that will have to be taken into account in the OSS. While the OSS is out of scope of ETSI NFV work today, it cannot be out of the scope of a service provider deploying NFV. One of the key tasks for providers that want to embark on an NFV transformation journey is to design a target OSS architecture, assess their current OSS capabilities, and build the bridge between the two. Clearly, not all the OSS/BSS functions will be affected by network functions virtualization. This leaves room for adopting a gradual OSS transformation, with implementation beginning from the most affected areas, such as service activation or fault management. Here, the gap analysis also becomes a fundamental step to assess which existing OSS platforms can be preserved to guarantee as smooth a transition as possible.

Leveraging NFV for OSS transformation

NFV introduces new demands for OSS and hence challenges during the transformation journey. Addressing these will result in a new layering of service and resource management that have implications across fulfillment and assurance.

NFV demands to OSS transformation

First, NFV, through the virtualization it introduces, brings new abstractions in the form of VNFs, Network Services, and other components—together with a new set of management components: NFV Orchestrator (NFVO), VNF Manager (VNFM), and Virtualized Infrastructure Manager (VIM). This will have a direct impact on the operational processes used in the OSS. Second, NFV requires increased automation. This will result in operations simplification, higher business agility, and faster time to market. It will, however, put added responsibility on OSS and NFV-MANO.

Finally, NFV technology brings new capabilities, enabling new business models for service providers. Examples include providing external network functions or services (from innovative service provider entrants) “as a Service” (FaaS), or simply using external infrastructure in

© European Telecommunication Standards Institute 2014. Further use, modification, copy and/or distribution are strictly prohibited

NFV introduces new demands for operations support systems. Adopting NFV means transforming the OSS, whether a radical change or a gradual approach is chosen. Addressing them will result in a paradigm shift toward services.

an IaaS model. Operations will need to focus much more strongly on customer experience management of their services rather than on managing infrastructure. The OSS needs to support such business models and enable such supplier or partner integrations.

With NFV driving virtualization, automation, new business models, and focus on customer experience, it becomes the catalyst to the business transformation. As a result, it catalyzes the OSS transformation. Through better customer experience management and understanding of customer behavior, OSS becomes a key factor in using and gaining maximum benefits from network functions virtualization, driving more flexibility, lower starting cost, and better monetization of services. To exploit its maximum benefits, NFV requires new thinking around the OSS that will offer opportunities to gain additional operational benefits. Depending on their individual strategy, some service providers will wish to evolve their OSS incrementally to accommodate NFV, while others will wish to exploit NFV introduction to make a step change in their systems. For those on the step change path, NFV could provide an opportunity to leverage NFV-MANO to transform the current OSS into a more efficient system. For those who wish to take the incremental path, it can be introduced in a way that minimizes the impact on existing OSS and operations models. In both scenarios, service providers may wish to leverage existing in-house IT skills to provide and manage NFV infrastructure, without requiring a wholesale change out of current network operations models and extensive retraining.

A new layering of service and resource management

TM Forum Frameworx is a suite of best practices and standards that is used by most OSS. Within Frameworx, the Information Framework (SID) defines the concepts of service and resource management. But with the introduction of NFV, which introduces the term “Network Service” and manages resources as virtual services, the meaning of the SID-defined terms becomes blurred. ETSI defines Network Service (NS) as a composition of network functions, defined by its functional and behavioral specifications. An NS includes VNFs and possibly Physical Network Function (PNF) endpoints. From a SID viewpoint, it can be considered as a kind of virtualized resource-facing service (RFS).

Similarly, the VNF can be considered as a kind of logical resource from a functional point of view. While services might include NS only based on VNFs, it is quite likely, especially in the beginning, that services will contain VNFs and PNFs, such as nonvirtualized functions. First, not all the network functions will be virtualized at the same time, making hybrid services more likely. Second, there will always be some nonvirtualized network functions at places where physical connectivity is needed, such as eNodeB or BRAS.

So hybrid services will be the most common occurrence. The OSS needs to be able to deal with them while preserving the agility and automation needed for NFV.

Product Customer

Product & commercial

Customer relationship management

Resource-facing

service Network service

Network service

Customer & Internal

Network service management Virtual network function Phys, network function (LR) Network function

Customer & Internal

Network function management Virtual machine IT, compute, storage, network Telco, compute, storage, network Infrastructure

Customer & Internal

Infrastructure management Resource-facing service Customer-facing service Service

Customer & Internal

Service management

6

A more detailed layering, based on the classic resource management layer, may help to map appropriate functional building blocks. The figure contains four layers (expanded from the two classic ones) below the customer relationship management layer:

• Service Management—In SID, the Aggregate Business Entities (ABE) of customer-facing

service (CFS) and RFS are representations of produced services. CFS and RFS are managed by the service provisioning and stored in the service inventory. They have an associated hierarchy of other CFS or RFS, network services or functions (network element) that need be provisioned in lower layers or which are the activation points of the service, for example, the Home Subscriber Server (HSS).

• Network Service Management—At this layer, the service can be decomposed in virtualized

and nonvirtualized RFSs. Virtualized RFSs can be mapped to Network Services as defined by ETSI. Of course, there is always the possibility that the service only contains VNF, but as explained above, hybrid services will be the most common. The TM Forum Business Process Framework and SID do not yet define management of Network Services but this is covered by NFV-MANO.

• Network Functions Management—In this layer, Figure 3 indicates the notions defined by

ETSI, such as PNF and VNF. The PNF represents standard logical or even physical resources, as defined by SID and corresponds to what exists today without virtualization.

• Infrastructure Management—The level of physical and virtual infrastructure. It can be an

internal data center but also could be an external data center, with a cloud-based offering. Typically, the virtual infrastructure management is provided by the data center provider.

NFV impact to fulfillment and assurance

We have seen in the previous section that virtualization is making the TM Forum Frameworx layers of service and resource management become blurred. We will look now at the impact of NFV on fulfillment and assurance processes of the OSS.

Compute and storage Network

Logical environment Resource

fulfillment assuranceResource

Resource lifecycle management VNF VNF … VNF

OSS

Compute and storage Network Service

fulfillment assuranceService

Service lifecycle management

Resource

fulfillment assuranceResource

Resource lifecycle management

OSS

Figure 4. Changes for OSS

For service fulfillment, it is necessary to provision services with the speed and accuracy

needed for NFV. Cloud-based approaches can help enable such automation by tying together the relevant systems and providing an efficient development environment to deliver benefits, such as:

• Zero-touch commissioning: This is the ability to install any device straight from the box,

• Flow-through provisioning: When an order for a service goes in or demands for service are

changing, flow-through provisioning must automatically propagate the needed changes down to the relevant network elements, whether traditional or VNF. NFV technology will not replace all existing legacy environments. Hence fulfilment solutions have to support near real-time automated provisioning of cross-technology components.

• Network optimization: As discussed above, service demands change based on customer

behavior and the need to maintain a positive quality of experience (QoE). Services may be turned up and down, require more or less of the available capacity, and adjust to network conditions. Automating optimization based on these changes maximizes efficient use of resources and drives positive QoE. The demand and QoE monitoring and analysis, done in the service assurance block, and a closed-loop feedback between the assurance and fulfilment domains is another essential change in the OSS landscape.

Service assurance needs to start even before a new service (whether based on NFV or a hybrid

of classic and NFV technologies) is ordered and received by a customer: the service provider has to ensure that the contracted SLA and promised QoE are being met. This requires collecting key service parameters, such as identification of the resource faults, packet loss, latency, and jitter. In complex distributed networks, setting up the data collection, monitoring the key performance indicators, and predicting the resource fault impact to the service, and even more to the customer, is challenging and time-consuming. The abstractions introduced by NFV further exasperate the challenges.

In addition, we have seen that to be able to achieve network optimization in an NFV environment, service monitoring needs to be closely tied to service fulfillment with a closed-loop feedback. Many current OSS cannot support this since the relationship between the fulfilment and assurance domains is relatively static, with a significant latency in synchronizing inventory information between the two domains.

This does not deliver the expected agility promised by NFV. Cloud technologies bring additional dimensions in monitoring. While the NFV technology seems to bring a promise of more centralized monitoring running in the data center, it raises the complexity with the need of complex diagnostics, real-time insight to service and QoE impact, and preventative fault isolation and elimination.

The assurance systems for meeting NFV needs have to ensure:

• Monitoring real-time, end-to-end service quality, taking into account availability,

performance, usability, security, and more—even up to the level of the individual subscriber service instance to ensure SLAs and QoE are met.

• Discovery of end-to-end service, which may be delivered on classic and virtualized

environments, and reconcile it with the services as provisioned and understood by fulfillment.

• Providing closed-loop, real-time feedback to fulfilment system about degradation or

reduced utilization to secure additional resources or releasing existing resources to match the service demand pattern and affinity rules.

The OSS business needs and challenges outlined above are not unique to NFV or new within the notion of OSS transformation. They exist already, strongly driven by the mobile services environments, where 3G and 4G technologies have opened new business models to over-the-top (OTT) providers. The focus on individual subscriber QoE and monetization of the services through information started several years ago. What NFV has done is to introduce a much more dynamic environment that stresses existing assurance OSSs, further accelerating the demand for OSS transformation. Without OSS transformation, the full NFV benefits will not be realized.

HP vision for OSS transformation

Ambitious business goals and new models of organization and processes will introduce additional challenges for the OSS, which in turn, will require a new OSS approach that supports higher agility and productivity, comparable to a production facility. In addition, the growing flood of data received from the network, further increased with the new service offerings enabled by NFV—for example, for machine to machine (M2M) or the Internet of Things (IoT)— will provide additional impact on OSS.

Ambitious business goals and new models of organization and processes resulting from NFV will introduce additional challenges for the OSS. This, in turn, will require a new OSS approach that breaks the existing silos and migrates to a “dynamic production environment.”

8

These deep industry transformations brought HP to adopt the idea of the service operations factory and to focus on applying Big Data analytics capability to the OSS to address the required transformation.

The use of Big Data technologies is even more relevant for NFV as its intrinsic agility and increased model abstraction and automation are further increasing the complexity of the network data collected. Traditional performance and fault management approaches won’t be able to cope with the complexity and dynamism introduced by NFV. In ETSI, NFV and Big Data are unrelated. We believe, however, they can and should be combined to build the OSS of the future.

The next sections illustrate how HP has started leveraging the service operations factory and OSS & NFV analytics to provide and experience new technologies for the OSS of the future, completing the HP NFV Director capabilities.

Service operations factory

Analysys Mason defines the service operations factory (see [AM SOF]) as all the technical functions necessary for a new customer onboarding and after-sales support. This encapsulates all of the technical knowledge of the services and network, providing to the outside world a simplified interface that takes a pure customer-oriented service view.

The service operations factory is very much like a manufacturing organization in the modern industrial world that reuses production facilities for multiple products and combines them flexibly. Similarly, the future of an effective and successful carrier management system lies on its capability to support the launch of a new service faster and seamlessly. It also can view and manage it on an end-to-end basis, hiding the infrastructural complexity to the outside world—whether the network operations center (NOC) operator or the customer. And similarly to a production facility, the new OSS are to be governed by a set of specific service key performance indicators (KPIs). The advent of NFV, with its need for elasticity and automation, enables another analogy with manufacturing factories. The service operations factory of the future, driven by customer demand for new service offerings, industrializes the service production, like a robot assembly line in manufacturing, making the provisioning of network services and functions (resources) a fully automated business.

Figure 5. The HP OSS Service Operations Factory

Customer CSR

Services view

Network view

Service

operations

factory

“Manage productivity”

“Show services”

Legacy infrastructure Data center infrastructure

“Think production”

Service onboarding request Service KPIs Service support information

Service Operations Center

Ser vice reque st Ser vice irregul ar inform ation Service operation factory KPI Common data

Standard production NFV production Onboarding services

operations

Support services operations

The HP Service Operations Factory combines fulfillment with assurance, coupled with a common data framework, and monitors production through service KPIs. This approach is unique in the marketplace as it takes a holistic view of operations instead of splitting it, based on organizational measures.

This transformation into a factory-oriented architecture or operations, which has already started, provides these foundations:

• Flexible offers—The production facility gives control to the buyer to compose products and

services flexibly and to adapt them as late as possible. The underlying resources that enable the service are hidden.

• Lean production—Lean production makes prudent use of resources, blocks them only when

necessary, makes them available as quickly as possible, and allows for late changes in the production process.

• Quality control—Continuous quality control of all production steps, with early identification

of bottlenecks and errors, reduces direct and indirect costs significantly while driving customer satisfaction.

• Continuous metrics—Standardized customer, service, production, and financial metrics

collected in near real time allow for continuous supervision of all production processes and alignment with contractual obligations.

For service providers, this means moving from technology-centric operations (typically structured in silos) into a service-centric operations environment. The entire end-to-end production process moves from a set of disconnected activities toward a mode where fulfilment and assurance operate in lockstep. Production control moves from hardwired activities toward flexible, data-driven control mechanisms. And finally, management moves from “flying blind” to a fully integrated, business-KPI driven mode of operations.

From To

Technology-centric operations Service-centric operations

Disconnected fulfillment vs. assurance Fulfillment and assurance in lockstep

Complex integration with long time to market Flexible service composition

Flying blind Fully integrated business KPI

The service operations factory concept enables these paradigms:

• One set of end-to-end processes—Siloed solutions lead to inefficiencies and disrupt operations.

Processes need to flow end-to-end across functional blocks, sharing the same data.

• One set of accurate data—In a near-real-time world (such as NFV), accurate data is the

foundation to all business operations and service innovation.

• One service model—A consistent enterprise data model of all services, applicable throughout

the service operations factory, ensures that fulfillment and assurance operate in lockstep.

• Preintegrated and accessible—Preintegration shortens time to market and lowers costs. A

services-oriented architecture enables easy access to functions in the OSS.

• Measure and supervise—Measurement and aggregation of all elements based on

standardized KPIs.

• Classic and NFV infrastructure—Standard and NFV-based production can be supported

and managed.

• Unified operator console—An overarching console provides unified access to all fulfillment

10

The figure depicts the major components and their interaction:

• The HP Service Operations Portal enables the operator, the customer representative, or

the planner to access the service onboarding (fulfillment) and service support (assurance) systems, offering them information with the right level of granularity for their job.

• Fulfillment provides the service onboarding and activation, triggering the assurance to

instantiate KPI measurement and SLA monitoring, and provides necessary data structures to enable correlation and analysis.

• Assurance provides the service support, triggers capabilities of fulfillment to automate

change and lifecycle management while supporting a more agile planning. It enables assurance to visualize network or service topologies.

• This is enabled since both act on the common data. While fulfillment provisions the data,

assurance is consumer of the data. Mastership of network data is the network. Reconciliation with the actual network status is essential for the common data domain.

The production infrastructure is managed by the service operations factory. Here we recognize the changes coming from NFV:

• The new NFV infrastructure is based on NFVI points of presence (PoPs) or grouped data

centers. The NFV Production combines the NFV infrastructure and the management of the

VNFs and Network Services.

• The existing classic infrastructure needs to keep its specific activation and domain

management systems, called Standard Production. In case the activation or monitoring is

completely independent from the virtualization layer—such as through an EMS or directly to device management interfaces—this Standard Production may apply to PNFs and VNFs. Production infrastructures can be provided by different suppliers for different domains. The factory idea enables hierarchies of subordinated factories, for example, to deliver access services, core services, or value-add services. ETSI defines the use cases to deliver NFV Infrastructure as a Service (IaaS) and VNF as a service (FaaS). Such a factory hierarchy is independent from the technology used (a single factory may use NFV or standard or both). Network functions virtualization can be the enabler that enables the transformation of a monolithic operation into a factory hierarchy.

HP supports specific solutions to manage IT or network infrastructures, specific to the domain (RAN, IP core, IMS, NFVI) or technology (classic and/or NFV). The generic architecture of the service operations factory can be applied to specific factory, but being implemented with specific solutions.

Fulfillment Assurance

Standard production

Common data

NFV production Service operations portal

Drive correlation and analysis End-to-end SLAs

Make planning guide Automate change and LCM

Visualize assurance Customers/CSRs/sales Self-service Self-care Service KPIs Operations/planning Service Care Figure 6. Components of Service Operations Factory and their interaction

In conclusion, the paradigm shift brought by the SOF is to combine the whole value chain (design, fulfillment and assurance) into a dynamic production environment that enables

the service provider to manage the service complexity required by the market, such as, a new service instance in minutes, not months.

Bringing together the two notions that we have described earlier of standard production and NFV production is the most natural initial instantiation of such a dynamic production environment. Such an implementation will be also greatly facilitated by the appropriate deployment of analytics capabilities and an adequately designed orchestration layer as we will explore in the next chapter.

OSS & NFV Analytics

Service providers have been using data warehouses for the last two decades to store the information coming from the network. The advent of Big Data, enabled by technologies such as Hadoop and HP HAVEn, has changed this picture, making it more affordable to store data and allowing better handling of its complexity. This evolution has made Big Data technologies more relevant to service providers in general and OSS in particular.

Operations is faced with complexity of today’s and tomorrow’s heterogeneous sprawl of environments and their produced data, as shown in the Figure 7. And all that in real time. Figure 7. Complexity of future’s heterogonous environmental sprawl

Monetize immediately Automate actionably Analyze anything Store everything Collect everything How do I know

what’s important? How do I know whenI’ll need the data? all this information?How am I to make of impact procatively and instantly?How do I reduce customer and increase business?How do I differentiate

Modern networks are providing more and more data and new technologies, such as Hadoop or other Big Data technologies, which provide the way to collect and store this data. The challenge is to analyze and automate response actions to minimize impact on customers and service. Transformation of OSS structures and solutions includes transformation of the technologies that cope with the mass of data the networks and their usage produces. It concerns fulfillment and assurance with their common data and mediation and, as seen in the previous section, is required to provide the KPIs needed for the supervision of the service operations factory. The complexity introduced by modern networks is further augmented by the introduction of NFV. NFV is bringing an increased level of model abstraction and decoupling, as well as an intrinsic agility at service and resource levels, that traditional fault and performance management approaches won’t be able to cope with. This is making analytics even more relevant for NFV. HP has been providing Big Data solutions to IT and service providers for some years. The company has matured these Big Data solutions in the telecom market to address customer experience and customer intelligence, with many of these solutions in production today. Moving forward, Big Data will further extend its footprint toward OSS and NFV. HP OSS & NFV Analytics is a program that consists of enabling new OSS and NFV offerings by leveraging Big Data technology. It answers new needs of our clients, namely the CSPs that want to take advantage of the enormous volumes of network and operations data, at an affordable cost, while gaining the necessary insight to derive greater business value from their investment. This approach enables them to further optimize their operations, better satisfy their subscribers, and monetize the intelligence buried in the data.

HP OSS & NFV Analytics will allow incorporating analytics directly into the NFV orchestration management and operations optimization processes.

12

As shown in Figure 8, the HP OSS & NFV Analytics program leverages Big Data techniques in the OSS and NFV space, streamlined with the mainstream HP New Style of IT Big Data, based on HP HAVEn.

Hadoop/HDFS

Autonomy IDOL

Vertica

Enterprise security

Operations Dashboards

OSS analytics foundation

Analytics for quality

of experience …

Analytics for predictive management

Solution consulting

and design

Deliver or run as a service

Business continuity

and support

Service and support

Applications

(CSP specfic analytic use cases)

CSP data sources

(Network and IT systems)

Market data

Subscriber data

Network data

Application data

Social media eMail Video Texts Audio Mobile probe dataNetwork Ticket data performance dataNetwork Networkfaults Images

n

Apps

Figure 8. HP OSS & NFV Analytics program

HP OSS & NFV Analytics program enables the following:

• High volume and variety of data can be analyzed cost-effectively in real time and overtime • Increase in data-loading speed

• Reduction in data compression, maintenance, and support cost

• Apply innovative analytics technologies to enhance their OSS capabilities • Reduced time to run queries and get real-time responses

• Easy implementation of any data model on any source, with no change needed and no additional response time with data growth

• Saving in storage space • Leverage commodity hardware

Data coming from the network will be collected, aggregated, and filtered, if needed. Information analysis will be applied on this data, and based on this analysis, potential problem will be detected, along with estimated ramifications on services, customers, and the business. As a result, preventive action will be taken in real time to forestall potential impacts. This will automatically lead to a stronger focus on service impacts and customer experience.

The same approach will be extended to an analysis of historical data to discover how predictive models can be refined and operations can be improved to move from reactive to proactive response. This will deliver better value in terms of capital expenses and operating expenses, and ultimately, in providing customer value and experience. Similar analysis on other data can be used to improve customer service or network engineering.

Currently, most of the OSS would provide automated responses for 20% of the problems discovered, letting operators handle the remaining 80%. One of the goals of OSS & NFV analytics is to reverse this ratio, providing automated response for 80% of the problem discovered, as well as a more accurate identification of problems, providing the level of automation needed for NFV operations.

This paradigm shift will use real-time, historical, and near-real time analytics. It will include the ability to collect any type of data from network resources, applications, operations, and relevant social media. HP OSS & NFV Analytics program enables comprehensive, actionable insight into all aspects of assurance and fulfillment.

While quality of experience will be the first focus of HP OSS & NFV Analytics, it will provide analytics solutions to all levels of management: NFVI and VNF management, classic resource management, end-to-end services, and customer management. As already stated, thanks to NFV technologies and the service operations factory approach, the production of a service is industrialized, comparable with modern, just-in-time car manufacturing principles. Combining this approach with HP OSS & NFV Analytics will further enhance it, providing the level of automation needed for reaching the benefits expected of NFV.

These new technologies and HP capabilities of industry experts and technologies, in tight interaction with our CSP clients, enable operations to become more successful and efficient.

HP NFV Director

HP NFV Director (see [HP NFVD]) is a key component of HP’s vision for OSS transformation. It handles the NFV Production part of the service operations factory. It provides a common point to ensure consistent management and behavior of VNFs, regardless of vendor, enabling each VNF to efficiently run on heterogeneous hardware platforms and virtualization environments. HP NFV Director takes responsibility for automatically managing the end-to-end service across VNF, VNF forwarding graphs (VNF-FGs), and network services (NSs).

It is designed to meet the evolving ETSI specifications for the NFV orchestrator functionality. This includes the orchestration and management of virtual network functions and network services, providing the global resource management, and consistently applying global, cross-VNF, and VNF-specific policies.

Within the ETSI model, VNF managers are responsible for the VNF lifecycle actions—deciding to scale in or out, for example, HP NFV Director can work with external VNF managers when supplied by the vendors. It also can provide the VNF manager functionality through its embedded functions, which compensate for completely or partially missing VNF manager functionality in the vendor solutions or in VNF created from basics by carriers.

Figure 9. NFV Director overview

Virtualized infrastructure managers/Wide-area infrastructure managers Global resource inventory Instance inventory Catalog Global resource

fulfillment Global resourcemonitoring Policy management

Service fulfillment Service monitoring

VNF fulfillment monitoringVNF Embedded VNF manager VNF fulfillment monitoringVNF External VNF manager HP NFV director

HP NFV Director is also responsible for the global resource management, allocating resources from an appropriate pool that satisfies the requirements of the VNF and of global policies (beyond the scope of a VNF). The global resource management policies can include provisions for disaster recovery, with separate disaster recovery processes automatically handling the recovery scenarios.

14

The following table summarizes key features of HP NFV Director:

Implementing NFV management in OSS

To complement the vision presented in the previous sections, it is important to be able to transform this vision in implementable steps.

HP has adapted its existing HP OSS blueprint to include the ETSI concepts for NFV Management and Orchestration. Although the ETSI ISG leaves the OSS out of its scope, there is no doubt that MANO components are technically an intrinsic part of the OSS.

We have seen in previous sections that most services, at least initially, will be hybrid services, containing virtualized and nonvirtualized network functions. This means there is standard production infrastructure, with a mature OSS in place to handle the standard resource provisioning. Not all service activations will need NFV management. Even for Network Services that are completely virtualized, there may be a need to handle the application-specific configuration of the various VNFs.

An example of such configuration is the creation of a VoIP user profile in an HSS, which is accessed through an element management system (EMS) or the network element management interface. In the ETSI NFV-MANO architecture, this configuration step is handled by the EMS, activated from the well-matured, standard production OSS.

NFV will first impact the resource and RFS layers while the CFS layer of service management will be less impacted, even though it needs to support the required service agility.

While a Greenfield approach to NFV is always possible in some limited cases, we have seen that the most common case will be a hybrid one, mixing virtualized and nonvirtualized resources. There are two different practical approaches possible to bring NFV into OSS.

Minimize the changes on the OSS

The first approach is to minimize the changes to the OSS by doing incremental integration in small steps. After the NFV production component is in place, it enables incremental integration in small steps, minimizing the changes to existing OSS. The service operations factory concept introduced in the Vision section enables decoupling of the higher level of services from the details of the resources and associated RFS. This enables an incremental transformation of the OSS.

Feature Description

Provisioning capabilities • Orchestration of VNFs, VNF forwarding graphs, and network services • Global resource management

• Works with external (VNF-supplied) VNF managers or directly provides VNF manager functionality • Seamless provisioning across physical network functions and devices in addition to VNF

Monitoring capabilities • Provides the required monitoring of the virtualization components

• Correlates physical infrastructure information into logical and virtual infrastructure information • Correlates VNF information into network service information

Multivendor • Multivendor orchestration across

• VNF, whether the VNF includes VNF manager or not

• VIM, including HP CloudSystem, Helion, OpenStack, and others • Hypervisors, including KVM and VMware

• Infrastructure (servers, storage, and networking)

Multisite, multitenant • Can orchestrate across multiple sites, managing placement of components based on requirements and policies • Intrinsically supports multiple tenants, whether the tenants are customers or operational domains

Supports multiple ETSI interaction models

• Provides a VIM proxy interface to the VNF Manager • Supports 3rd-party NFVO-VNFM-VIM interactions

Accepting this approach, we will describe the new OSS, using three major groups of functional building blocks:

• An overarching End-to-End (E2E) Service Management provides end-to-end views of the

service, independently from traditional (standard) or NFV infrastructure but keeping the knowledge, if network services are based on NFV or standard, or combination of both. Most service catalogs can already define this, with the service orchestration directing the order to the lower level OSSs. We call these Standard Production and NFV Production.

• The Standard Production component will use existing OSS to perform resource provisioning

of PNFs or to activate services independent from the NFV means, such as through the EMS for the VoIP user profile on the HSS. If the service is decomposed into standard RFS and Network Service, Standard Production will be triggered by the E2E Service Management. It might also be triggered from the NFV Production, for example, to configure a PNF endpoint.

• The NFV Production component consists of the ETSI NFV-MANO building blocks: NFV Orchestration,

VNF Manager, and Virtualized Infrastructure Management (VIM). The NFV Production may actually use Standard Production when a Network Service comprises PNF endpoints. Such PNF may require complete resource provisioning or just configuration through the PNF’s element manager or management interface—both possibly using the Standard Production systems. The data mastered by the E2E Service Management are customer-facing services with associated resource-facing services. The CFSs are not really impacted, whether they rely on virtualized RFSs (Network Services) or nonvirtualized RFSs, as the services they provide to the customer are not intrinsically changing with NFV. The RFS will be associated with network and standard services, including their VNFs and PNFs, which are mastered by the lower level standard or NFV production.It presumes the E2E Service Management layer and the Standard Production have the right characteristics, such as agility and automation as described earlier, to meet the needs of NFV management and can evolve for the NFV promises to be fully realized. This approach is best for limited VNF introduction that won’t justify significant OSS changes.

Leverage NFV as a catalyst for OSS Transformation

We know, however, that a number of existing OSSs don’t exhibit the required characteristics, and evolving them might be too costly. The second approach proposed is to leverage NFV as a catalyst for OSS transformation. In this case, concepts seen in the HP NFV Director section would be used as the starting point of a modernized OSS.

Standard

production productionNFV

PNF PNF EMS VNF VNF

Classic infrastructure NFV-infrastructure

Customer Planning Operator Engineer

BSS

OSS

Production

network

Internal order for network service

Standard request NFV request

Standard request Customer order for

network service Customer order foruser service

E2E Service Management

16

Using some of NFV-MANO concepts, the E2E Service Management layer is extended to handle hybrid service modeling and orchestration, providing the automation and agility needed for NFV. Being able to manage in a closed loop and automated manner, hybrid services will be a key requirement for this E2E Service Management layer.

Depending on the exact needs, the resource management layer can also be extended with NFV-MANO concepts to cover the overall resource management for virtualized and nonvirtualized resources, leaving the standard production only the basic provisioning and interfacing with the EMSs. This is shown in Figure 11:

NFV production Standard production Standard production PNF PNF EMS VNF VNF

Classic infrastructure NFV-infrastructure

Customer Planning Operator Engineer

BSS

OSS

Production

network

Internal order for network service

Standard request NFV request

Standard request Customer order for

network service Customer order foruser service

E2E Service Management E2E Service Management

Figure 11. Extending NFV concepts

The default recommended option to the first option is to start incrementally with small steps. This enables a gradual introduction of NFV and a smooth transition of the operational

management. Where existing OSS cannot accommodate this option, or if a faster transformation to NFV is desired, a more holistic approach of OSS transformation is recommended. Choosing the best approach is an individual service provider choice, where a consulting approach might help.

HP Solution Consulting Services for your transformation

Approaching OSS adaptation to NFV is a challenging but significant opportunity for service providers to better integrate, standardize, and optimize their operational model. It removes traditional barriers, especially among operational readiness, fulfillment, and assurance. While the operational model is typically vertically designed to grant the proper level of first and second level of support, adapting it to a real-time service activation and on-demand capacity management will require a more flexible organization—and at the same time, end-to-end service lifecycle capacity to properly cope with the increased agility demand.

As a consequence, transformation governance in OSS is at the core of successful transition to NFV as consistently addressing the different dimensions of process, organization, and technology. We also have the firm belief that every client needs to approach that transition in its own way. A proven approach that properly shapes the new operational model, aligned with business goals, and driving the technology changes is key to successfully manage transition. HP Solution Consulting Services helps providers shape the strategy by delivering such a structured approach. We work together with the customer influencers and stakeholders, combining the business and the technology drivers in the proper way to ensure success.

Conclusions

NFV is a game changer in the communications industry. It promises new business goals, reduction of CAPEX and OPEX, acceleration of innovation, and improved time to market. NFV is disruptive, however, and introduces a new set of essential challenges to operations and the OSS. These challenges introduce new demands for OSS and make NFV a compelling event to drive OSS transformation, as management is a critical success factor for NFV introduction. HP’s vision for OSS transformation leverages the service operations factory and OSS & NFV analytics to provide and experience new technologies for the OSS of the future, completing the NFV-MANO capabilities provided by HP NFV Director. It addresses the paradigm shift toward service-centric operations, providing the agility, flexibility, and automation needed to make NFV successful. It also enables a smooth transition and seamless integration of the new NFV management into existing OSS/BSS structures.

HP has a complete suite of products and solutions to implement the vision of the NFV-ready OSS. HP can provide products for all functional building blocks, letting the service providers choose the implementation path best suited for their organizations and businesses. To complement this offering, HP CMS Solution Consulting Services provides concepts and guidance for the transformation into a successful NFV business case. This combination is unique in the industry, making HP the best open solution for OSS and NFV management.

Learn more at

hp.com/go/oss

hp.com/go/nfv

Rate this document Share with colleagues

Sign up for updates

hp.com/go/getupdated

This is an HP Indigo digital print.

Acronyms

ABE Aggregate Business Entities

BSS business support systems

CFS customer-facing Service

CSP communications service provider

EMS element management system

ETSI European Telecommunications Standards Institute

FaaS VNF as a Service

IaaS Infrastructure as a Service

IoT Internet of Things

ISG Industry Specification Group

KPI key performance indicator

M2M machine to machine

NEP network equipment providers

NFV network functions virtualization

NFVI NFV Infrastructure

NFV-MANO NFV Management and Orchestration

NFVO NFV Orchestrator

NOC network operations center

NS Network Service

OSS operations support systems

OTT over the top

PNF Physical Network Function

PoP points of presence

QoE quality of experience

QoS quality of service

RFS resource-facing service

SOF service operations factory

VIM Virtualized Infrastructure Manager

VNF Virtual Network Function

VNFM VNF Manager

References AM SOF

“Towards a services-oriented factory super-suite”, Analysys Mason, May 2013

http://www.analysysmason.com/

HP NFVD

“Realizing virtualization, HP delivers network functions virtualization orchestration”, HP NFV Director Solution overview, Hewlett-Packard, February 2014, Realizing virtualization

© Copyright 2015 Hewlett-Packard Development Company, L.P. The information contained herein is subject to change without notice. The only warranties for HP products and services are set forth in the express warranty statements accompanying such products and services. Nothing herein should be construed as constituting an additional warranty. HP shall not be liable for technical or editorial errors or omissions contained herein. 4AA5-6431ENW, January 2015