DC Cloudification & Consolidation Smooths Digital Transformation for Operators

By Wang Wei, Data Center Integration Solution Domain General Manager, Huawei

The cloud transformation that Huawei helped operator W implement comprised four steps: The first was forming a hybrid cloud blueprint and planning the migration rollout roadmap for the solution with the best return on investment (ROI). The second was building the cloud platform, consolidating the ecosystem, and putting everything into an ‘as-a-service’ model. The third is a smooth move to the cloud for all cores services while significantly reducing OPEX. The fourth was continuous strengthening and optimization of the cloud environment, improving O&M efficiency.

Goals of the transformation included full adoption of the ‘as-a-service’ model in the new IT architecture; adoption of a unified management and O&M of multiple private clouds and public clouds; building an open cloud platform able to support growth with third-party partners and support coexistence of traditional, cloud-based and cloud-native services with full data and capability sharing utilities.

DC cloudification & consolidation often takes place during the digital transformation process of enterprises and it is also one of the paths leading to digital transformation.

The transformation of digital business is characterized by on-demand promotion, real-time business, full online, self-service, and socialization. This requires infrastructure to connect enterprises and individuals in a more efficient, flexible and open manner, and it is also more dependent on IT and network. The full synergy of the business, with the full support of new technologies such as cloud, automation, artificial intelligence, and big data analytics.

Good practice: Transformation process at mobile operator W

Operator W is the third largest mobile operator in its country in the European market. As the revenue of traditional telecom services slowed down and market competition intensified in the whole European market, income from W’s mobile services declined. The impact of a series of external and internal woes made the operator realize the ‘good old days’ of making a good return on traditional businesses and models were gone. They needed a transformative breakthrough, thus the path to cloud-based everything.

For W, the external woes included not only the Matthew effect from the accumulated advantages of the first-place operator, the even more pressing conflict was coming from OTT services. According to a report from OVUM, Instant Messaging applications provided by OTT providers will reduce the SMS service revenue of European operators by 34% from 2015 to 2018 and by 2020, OTT VoIP will bring a total loss of 73 billion USD to telecom operators.

Internal woes laid in the aging IT infrastructure, slowed response to customer requirements, and lack of support for business development.The IT system was made up of a wide mix of appliances, legacy systems, and so on.

Virtualization was applied to some degree but the cloudification process had not yet begun. This meant that elastic scale-out of the CRM and billing systems was not supported. Adding to the headaches, orchestrating applications was rather complex and there was no unified management platform to run all the cross-domain IT applications. New service rollouts were slow, and even adding functions was a painstaking endeavor with the old structure. For example, the BOSS system did not support flexibility in customizing service packages, there was a lack of online services and other functions for customers resulting in the low level of automation. Deployment of new services required intricate collaboration between five or more departments and even simple things like registering a DNS took two or more days to complete. Another huge issue was the extremely high OPEX. The cost to maintain IT equipment ran up to 20 million euros each year. The high costs in turn took away from the ability to re-invest in the IT infrastructure. For example, a large number of aging devices (mainly UNIX servers and storage devices) that had been running on the live network for eight to twelve years creating huge maintenance costs. Compounding the situation, it was also unclear which specific server was carrying services for around 75% applications on the live network. The list of issues continued into disarray at the equipment room with the dispersed and aging devices, and utilization of the in-place devices was extremely low.

W realized it had to make the transformation to the cloud to stay at the top of the game.

Adding to the depth of the required consolidation, W merged with H, the fourth ranked operator in the local market. The move would help the operator gain share in the local market, improve competitiveness, and accelerate the breadth and depth of the consolidation in the data center layout. The project had two core objectives. The first, quickly get OPEX under control and relieve some of the operational pressures by shutting down and consolidating the aging data centers while also replacing UNIX with x86 servers and migrating traditional applications to the cloud. The second was to build an agile IT mechanism capable of unified architecture management, fully implementing cloud-based as-a-service capabilities, and optimizing the approach to O&M to accelerate service innovations and better respond to the challenges in the digital transformation.

To quickly reduce OPEX and accelerate service innovation, Huawei helped W break the DC cloudification & consolidation project into four major steps. Each step has an important goal:

Step 1: Formulate the hybrid cloud transformation blueprint for the entire group and analyze the roadmap to find which solution offers the best ROI.

Step 2: Build a software-defined data center (SDDC) to simplify IT management and deliver a cloud-enabled, service-oriented infrastructure.

Step 3: Migrate core services to the cloud smoothly and achieve quick reduction in OPEX.

Step 4: Place services from the hybrid cloud environment to the deployment end online and continuously improve management practices.

First step: Formulate the hybrid cloud blueprint and select the solution with the best ROI

W planned to make a strong push into B2B services over the next three to five years and establish a new business department to develop services such as online shopping malls and digital channels. The new department started using a public cloud and cloud native architecture to accelerate service rollout and improve platform flexibility and reliability. Considering that such core applications as BSS and OSS would need to remain in the traditional IT architecture for a considerable period, Huawei helped W design a SDDC reference architecture able to support coexistence of the IT Bimodal (dual-mode IT architecture). Huawei also helped formulate a five-year phased implementation roadmap in moving its applications to the cloud. The plan considered the service priorities and expected benefits with precise in-event execution for the various layers of W’s core services.

Huawei’s hybrid cloud transformation consulting service provides end-to-end consulting services for everything from cloud data center planning, service continuity and security planning to evaluation of the cloud adoption paths for applications and the organizational processes in O&M operations. This service provides customers with a transformation blueprint, reference solution architecture, and roadmap. Huawei has become the first choice for many customers undertaking a cloud transformation.

Second step: Build the platform, integrate the ecosystem, and make all services available in ‘XaaS’

The BSS, OSS, and MSS domains were ran as independent systems in the W’s operations. The virtualized resource pools comprised HP, IBM, VMware, Microsoft, and Redhat products. These resources were managed by different departments and each was deployed on a different resource to try to match the service characteristics of the specific business, making management extremely complex. This approach meant many resources could not be shared amongst the varying services because there was no unified approach in managing the resources. Resources were overwhelmed during peak usage times while others remained almost completely idle. To address this situation, Huawei helped the operator plan a unified hybrid cloud management and O&M platform for the entire network. First, centralized management capability over the five heterogeneous virtualization resource pools and the physical machines on the live network was enabled to simplify the IT framework. Second, the public and private cloud built by two vendors were integrated to build a unified portal for using, monitoring, and managing the profiles.

Third, through unified service orchestration and automation, 40 types of common application templates for various departments were produced. These templates freed up administrators from having to perform complicated service application, approval, installation, provisioning, configuration, and provisioning processes. Fourth, the IT capabilities of the live network were placed into cloud containers and the IT capabilities of three commonly used partners were placed into the XaaS catalog, accelerating service innovations. The resources in the heterogeneous mashup are now combined to deliver computing, security, disaster recovery, and monitoring services on demand.

The Huawei cloud infrastructure integration solution provides service-driven resource pool design and a unified cloud management platform able to integrate existing resources and enable unified service orchestration, unified monitoring, and cost visualization for multi-cloud and hybrid cloud profiles. The IT capabilities of the live network and those provided from partners were placed into a XaaS cloud profile to provide a rich set of flexible SLA assurances for upper-layer services.

Step 3: Smooth migration to the cloud, quick reduction in OPEX

W planned to gradually migrate the 400+ applications on the live network to the cloud within three years. A large number of aging and out-of-warranty midrange computers and storage devices that carried core application databases and middleware needed to be evaluated to determine which applications on the live network can be moved to the cloud. The operator had to consider which path and solution would best suit its needs while reducing service risks. The core applications running on the live network were provided by nine different vendors. After decades of build-out, the application dependency became complicated and the upgrade downtime and migration windows differed considerably between services where overlapping migrations often lead to prolonged downtimes. The huge mash-up in software and hardware models at the bottom-layer made the migration event complex and the risk high. Therefore, before planning the migration solution, the eTOM model had to be considered in addition to the level of importance of the service and the benefit in the move to the cloud. Huawei cooperated with managed service company A to sort out the relationships between applications and the IT infrastructure. They used tools to collect information, conducted manual surveys and expert reviews, and divided the migration event in to groups and batches. The result from Huawei was a design that included over 100 migration and risk mitigation utilities. The actual events were conducted after ample test runs and verifications in the lab.

Huawei Migration Service is capable of migrating midrange computers, storage devices, databases, and applications. Complemented by the Huawei CloudMount migration tool chain, AI technology is used to collect massive amounts network data, accurately restore application associations, and implement intelligent batch planning and automated migration processes – all of which are visible and manageable throughout the entire process while ensuring the shortest service interruption time and zero data loss.

Step 4: Continuous optimization of the cloud environment and improved O&M efficiency

If a service is planned to be released on the cloud, 24 steps were required in the planning, resource preparation, service release, and monitoring phases in the previous layout. Half of the steps needed to be manually processed, especially the O&M, configuration change, and troubleshooting processes. The manual aspects of the processes are error-prone and inefficient. To address this issue, Huawei helped W sort out more than 40 common service use case from the perspective of the service life cycle and compiled more than 20 atomic scripts (including installation, check, and configuration scripts) that are orchestrated and managed by a set of automated process tools. The case and script library are continuously enriched from actual operational events.

Huawei Cloud Continuous Optimization Service provides performance evaluation and optimization suggestions at the resource and application layers as well as for service health evaluations and security hardening. With the Huawei OpsCloud continuous optimization platform, processes and scripts are orchestrated and executed in a unified Orchestration, simplifying O&M and improving efficiency

Huawei set three objectives in the transformation of IT architecture. First, enablement of unified management and O&M of multiple private and public clouds. Second, delivery of open cloud platform able to grow with third-party partners and continuously optimize cloud-based O&M operations. Third, support the coexistence of traditional services and cloud-based/cloud-native services with full data and capability sharing utilities.

Using Huawei ShapeCloud to design the Hybrid IT DC Cloudification & Consolidation Solution was the first step of the consultation in the hybrid cloud transformation. It provides customers with a distinct roadmap for hybrid cloud, a complete IT Bimodal blueprint, and a clear ROI. The second step was building an OpenStack-based open cloud platform to centrally manage traditional and cloud-based IT appliances and applications and place third-party partner IT capabilities into the cloud service package. New service rollouts were shortened from 6 to 8 months down to 2 to 4 weeks, rapidly accelerating ICT service innovations oriented to enterprise customers. Step three was replacing aging devices and migrating services to quickly reduce OPEX (42% reduction in TCO anticipated over five years). The last step includes continuously optimization of services to improve O&M efficiency and business performance. The staged build culminated in a service-oriented and simplified cloud data center layout that is helping the operator achieve the core objectives in the transformation of the IT architecture

Digital transformations only have a starting point, no end point. Find out more about how Huawei can help you plan your digital transformation.