Welcome to the cloud: Business transformation through next-gen cloud evolution

2016.06.02 By Allan Mow, Gracen Duffield & Lu Xi

Burdened by a rigid infrastructure, telcos are struggling with high total cost of ownership (TCO), exacerbated by a lack of agility, visibility, and mature systems. Under threat from agile competition and with income shifting away from voice and traditional data, telcos are under great pressure to maximize efficiency and create revenue streams by delivering new customer-oriented services. 

Unsurprisingly, more telcos are looking to cloud evolution to help remove pain points and get up to speed with industry trends. By building next-gen service-driven cloud data centers (CDCs) based on open cloud platforms such as OpenStack, they can reduce costs, become more innovative, and accelerate time to market (TTM) to overcome challenges and stay ahead of the game.

In the clouds

High TCO

Traditional IT costs eat up as much as 8 percent of an operator’s gross revenues in developed markets (Feldon 2014). The primary contributors are low resource utilization, vendor lock-in, and aging equipment. 

Due to siloed infrastructure, we’ve found that CPU utilization for non-virtualized servers averages between 10 percent and 20 percent, with peak utilization rarely topping 60 percent. Often designed to accommodate a handful of once-a-year Black Friday-type peak load events, excess capacity sits idle for the other 360 or so days. One Huawei customer found that average CPU utilization for one of its opcos was a pretty dreadful 16 percent. Even worse, it couldn’t deploy new applications due to a lack of available resources. 

Locked-in to proprietary technology places telcos at a significant disadvantage when it comes to negotiating with vendors on discounts on licenses and hardware renewal. In a more subtle form of lock-in, vendors might only support certain certified configurations. When 70 percent of its servers reached the end of its support lifecycle, one Huawei customer discovered that licensing and configuration restrictions stopped it moving applications to available servers, incurring a whopping US$200,000 bill for new licenses – twice as much as an OpenStack solution. The final nail in the coffin is older equipment, which uses more space and power, is inefficient, and raises costs when it begins to wear out.

Through consolidation and by virtualizing server and storage solutions, open-cloud architecture provides shared resources in consolidated frameworks under automated unified management systems. Open technology protocols with multiple hypervisors integrate heterogeneous commodity equipment. These improvements drive down TCO by helping telcos improve resource utilization. They also evade vendor lock-in and assess older equipment TCO to reduce power, cooling, and facility costs. CDCs also provide better management tools for automation, lowering labor costs on low-value tasks.

Lack of agility

With traditional siloed infrastructure, segregated resources cannot be shared and deployed when they’re needed, instead sitting idle. Complex application integration coupled with a mish-mash of software platforms makes adopting a shared service model for efficiency and cost saving extremely difficult, because supporting each system requires different skills and knowledge. This leads to capacity constraints that hinder IT delivery as each new project requires an additional buildout when DCs are at near-full physical capacity. Meanwhile, the procurement and installation of the dedicated infrastructure needed for each new project extends lead time. 

Cloud infrastructure eliminates the need for clunky procurement processes, and makes it possible to deploy shared heterogeneous resources and manage them intelligently on a shared platform. These improvements boost agility and TTM, which can in turn support the development and competitive delivery of diverse digital and cloud services with minimal lead times.

Lack of visibility

Problems with high TCO and a lack of agility are worsened by poor visibility when it comes to capacity, performance management, and SLA attainment due to the lack of standardized architecture and processes. 

With opcos often operating independently or acquired through M&As, telcos lack a consolidated, real-time view of their business-wide resources because different tools, reporting structures, and management portals are in play. One Huawei customer, for example, was unable to produce consolidated reports on the capacity of each of its infrastructure domains and application layers because its server, storage, backup, and network teams used different data capture tools. Without a working CMBD, it was unable to correlate raw data and services.

Traditional telcos also lack end-to-end performance management tools, making it difficult to quickly identify the cause of performance issues, despite the use of reactive monitoring to alert support staff about compromised assets. Meanwhile, telco IT departments often focus on specific task completion metrics in their SLA management and KPIs, rather than correlating performance with financial measurements, which telco executives care more about.

Cloud transformation can provide better visibility on current capacity, performance, and KPI levels to provide better support for decision making and integration with partners and suppliers.

Growing up

Telcos often find themselves tied down by their existing processes, tools, and technologies. For instance, manual processes, workflow tracking, and poor change management can be destabilizing and impact service availability. Too often, DR capabilities are untested, which can impact availability and SLAs. IT directors often feel their company would struggle with a disaster-level event because the backup of siloed applications is extremely inefficient. SLA management is not always automated, often focusing on captured results, not proactive management. 

Service requirements must be checked against current availability and DR capabilities. Cloud transformation helps an organization mature by virtue of better IT management control, processes, and operations.

An overview of the cloud transformation process

It all begins with strategy – assessing industry trends, opportunities, and pain points; understanding customer goals; and defining an action plan. This is followed by Discovery and Analysis, where gaps and requirements are identified by comparing current and end-state IT. TCO is analyzed, and people, processes, and technology are assessed to ensure strategies can be fulfilled. Then Planning and Design sets out each stage of transformation, including implementation, execution runs books, and migration waves. During each implementation stage, new cloud architecture is installed, tested, validated, and migrated, leading to knowledge transfer. Finally, continuous improvement involves validating services, optimizing, and providing guidance to ensure continued excellence.

Implementation: four steps to the cloud

Implementing cloud architecture is a four-stage process. Stage one involves server and storage virtualization to increase machine efficiency and reduce complexity. This is followed by resource pool architecture construction in stage two, which helps build operators’ capabilities for innovating IT services and speeding up TTM. In the third stage, unified management tools, performance monitoring, capacity monitoring, and SLAs are put in place to enable unified management and automation. Stage four sees a flexible and cost-effective cloudified architecture provide service virtualization and consolidation.

During the cloud evolution process, telcos must implement mature processes and operations that span day-to-day operations, ITIL process frameworks, and technology domain readiness. By implementing practical improvement plans, operators can be sure that their people, processes, and tech can support cloud transformation and deliver business returns.

Transformation deploys technology and processes that support a broad range of business goals. Among these, telcos often first choose an internal cloud to optimize their OSS and BSS alongside new big data applications. XaaS, including IaaS, can then be developed for external customers. As sophistication rises, NFV (network functions virtualization) and network support can be used to create dynamic network functions for next-gen customer cloud services.

What does the end picture look like?

With a distributed cloud architecture and point-to-point interconnections, operators can achieve unified management, presentation, and operations. Global operators‘ CDCs in different regions allows scalable logical resource pools that incorporate physical and virtual resources.

To follow industry trends, telcos need a unified, consistent, and long-term vision of a comprehensive, inclusive, agile, and unified cloud architecture: "Comprehensive" satisfies current and future CT and IT requirements; "inclusive", or open, supports low-cost heterogeneous COTS (commercial off-the-shelf) hardware and software; and "agile" deploys applications and services faster, lowering TTM.

OpenStack and legacy support

According to Gartner, 75 percent of operators employ a dual-hypervisor strategy for virtualization and cloud. They do this to balance risk management, avoid vendor lock-in, and save costs. But, it complicates their business and, in the long run, increases OPEX. Deploying an open cloud platform such as OpenStack over multiple hypervisors unifies management and saves costs. 

OpenStack also provides flexibility for future innovation and cuts costs by 30 percent compared to proprietary platforms.

First migrating less critical applications to the open cloud platform can minimize risks and immediately cut costs – it also gives IT staff experience using the platform while they develop the ecosystem.

In a nutshell

More than just a new piece of hardware, cloud transformation also matures and modernizes operations by integration and migration, and increases utilization through resource consolidation and intelligent virtualization. 

By transforming IT, telcos both reduce costs and expand their capabilities to develop new offerings. According to Gartner and Huawei Market Intelligence, transformation lowers TCO on existing infrastructure from an average of 60 percent to 48 percent, while almost doubling spending on innovation from 19 percent to 35 percent. Overall, cloud transformation cuts IT costs by an impressive 20 percent. 

In a dynamic environment replete with challenges, cloud transformation helps operators overcome chronic problems, frees up resources to focus on new objectives, and cultivate opportunities to ensure long-term business success.

qrcode comm88 en

Mobile reading