Size daha iyi bir deneyim sunmak için çerezleri kullanıyoruz. Ziyaretinizi varsayılan ayarlarla gerçekleştirerek, çerez kullanımına izin vermiş oluyorsunuz. Çerez politikasını inceleyin.
Emir Halilovic2020-10-26 13
Expansion of enterprise data centers brings forward requirements for high-capacity, high-quality, flexible, and scalable connectivity. Driven by these requirements, enterprises across different verticals are migrating toward self-built and self-managed Data Center Interconnect (DCI) solutions.
The needs of enterprises, however, additionally dictate other features of evolving DCI solutions. These include simple and automated operations, administration, and management as well as network architecture simplification.
In this blog post, GlobalData’s Emir Halilovic considers the changes to the enterprise connectivity ecosystem, which are driving the adoption of DCI solutions.
As enterprise reliance on cloud services grows, several verticals have developed the need for large-scale data centers, generating significant volumes of traffic. At the same time, the availability and performance of DCI optical solutions are steadily growing. In line with these developments, building owned and self-managed DCI solutions using DCI platforms has become a viable and cost-effective alternative for enterprises seeking to satisfy their data center connectivity requirements, thereby sidestepping the downsides related to telco-provided connectivity services
Abundant Capacity: Data center connectivity needs to grow in line with the amount of data stored and processed within them, the number of users, and more complex applications requiring higher per-instance data flows. At the same time, more complex cloud infrastructures, covering multiple geographic locations with features like Disaster Recovery (DR) and backups, require very high-capacity and preferably low-latency links between data center locations. In many global markets, depending on the size of cloud infrastructures and the enterprises operating them, satisfying the need for abundant capacity by acquiring managed operator services can quickly become complicated and very costly.
Security and Compliance: Across different verticals, many entities — including financial institutions, governments, and utilities — are required to ensure high security of their data infrastructures, while satisfying stringent compliance guidelines for data security. For those organizations, the ability to ensure and manage the security of their communications on their own infrastructure is a highly attractive alternative to relying on operators to fulfill these requirements.
Scalability and Manageability: Changing needs of internal and external users, growing data traffic, and traffic pattern changes often require enterprises to acquire additional connectivity services 'on the fly.' However, many telecom service providers find it hard to quickly satisfy change requests such as this. Moreover, beyond stipulating Service Level Agreement (SLA) parameters, enterprises often have very limited control of service parameters and have almost no control over them once the service is installed.
These trends all favor self-built and self-managed connectivity solutions, which offer superior economies of scale. This type of connectivity can be built using high-capacity, simplified DCI optical platforms, operating on leased dark fiber or in conjunction with wavelength or spectrum services that operators increasingly offer.
To satisfy the needs of enterprise cloud infrastructures, DCI platforms should ideally have several key features.
Cost Advantage: Compared to Carrier Ethernet (CE), Multi-Protocol Label Switching (MPLS), or leased line connectivity services, self-built and self-managed DCI solutions include an upfront Capital Expenditure (CAPEX) component. However, the cost of self-built DCI connections does not increase linearly with growing capacity. DCI systems ideally should feature scalable throughput capacity, which can be increased by adding new modules and pluggable client and line interfaces. Lighting additional capacity in already installed optical systems featuring a high degree of modularity is much less costly than scaling capacity by acquiring new connections from the service provider.
Practically Unlimited Scalability: DCI optical systems in use today have capacity per optical fiber in the tens of Tbit/s. Additionally, due to the relatively simple process of adding new capacity on already installed and lit fiber, self-built DCI offers a high degree of flexibility for enterprises facing an increase in traffic, new application demands, and unexpected changing traffic patterns. Leading systems in use today allow operators to use the maximum optical bandwidth of the fiber plant they have acquired or built , using extended C and C + L bands, or even Super C and Super C + L bands.
Encryption, Quality, and Reliability: DCI platforms usually provide increased security compared to telco-provided connectivity services, by offering wire-speed optical encryption of their line interfaces. Additionally, leading platforms offer client encryption as well, allowing users to deploy End to End (E2E) secure capacity between any two points in their network. Control over the E2E DCI solution also provides the ability to achieve the highest possible quality of connection on any link in the network by controlling the performance of every network element involved. Finally, for critical links, leading DCI platforms offer established optical communication protection measures, ensuring superior reliability.
Flexible Deployment and Simplified O&M: The engineers in enterprise usually have Information Technology (IT) backgrounds; therefore, DCI Operations and Maintenance (O&M) systems need to be very efficient and easy to operate. DCI solutions deployed in enterprise environments should be developed with the aim of automating and simplifying O&M.
The evolution of enterprise cloud infrastructures also opens the path toward the transformation of enterprise data center connectivity. Performed in the right way, this transformation will bring enterprises a whole host of new capabilities, while providing the potential for significant cost savings as well as increased efficiency and productivity.
Huawei updated the capability of its DCI platform — OptiXtrans DC908 — to reach 800G per wavelength and 6.4T/U at HUAWEI CONNECT 2020 in Shanghai, further driving innovation in the highly contested DCI market.
Huawei OptiXtrans DC908’s updated features are still leading the market in terms of capacity and density of the platform, maximum capacity of the fiber, and automatic commissioning and intelligent O&M functionalities, which simplify O&M for data center engineers who mainly have Internet Protocol (IP)/IT backgrounds. All of these features make Huawei OptiXtrans DC908 attractive in all the industries it is used in.
Released in 2019, the first iteration of Huawei OptiXtrans DC908 offered:
• A rack-mountable 2 Rack Unit (RU) form factor with a depth of under 600 mm, making it suitable for deployment in both IT and European Telecommunications Standards Institute (ETSI) racks.
• Support for both a high-capacity discrete coherent solution as well as cost- and power-optimized CFP2-DCO pluggables on the line side.
• Class-leading power consumption of 0.13W/Gbit.
• Support for Fibre Channel (FC8G/12G/16G/32G) and Ethernet (10 GE/25 GE/40 GE/100 GE/400 GE) on the client side.
• A combination of electrical and optical modules on the same platform, making OptiXtrans DC908 efficient in smaller deployments by removing the need for a separate optical line system.
• The choice of an Artificial Intelligence (AI)-powered Network Cloud Engine (NCE) management and control system or a simpler element management system, depending on customer preferences.
In addition, updates made to Huawei OptiXtrans DC908 offer several technical characteristics that make it a compelling choice for DCI applications, including:
• Per-RU throughput capacity of 6.4 Tbit/s, matching the density of leading platforms in the marketplace.
• Support for maximum wavelength speeds of up to 800 Gbit/s.
• Maximum 88 Tbit/s capacity per fiber pair using Super C-band plus L-band.
• The WebGUI management system is able to support network dimension O&M. This is an ideal solution for small DCI networks and saves investment on the Network Management System (NMS) server.
• One-click automatic commissioning, from scratch to completion in minutes, automatically completes device discovery, fiber connection discovery, wavelength configuration, optical layer commissioning, service auto-adaptation, and protecting configuration.
• Intelligent, proactive O&M with an AI-powered NCE management and control system with numerous useful use cases.
Meanwhile, fiber fault prediction enables proactive optical fiber O&M. 20–50% of fiber fault is from gradual degradations, which are hard to locate in advance with engineers left passively reacting to service interruptions. With Huawei OptiXtrans DC908 and NCE, gradual degradation of fiber can be predicted in advance, with a 95% accuracy rate.
Co-bundling fiber-cable detection identifies whether fibers are in the same cable or route. This is helpful for protected DCI scenarios where working and protecting fibers should be in different routes.
Accurate root cause analysis simplifies maintenance to achieve one alarm per fault. The Dense Wavelength Division Multiplexing (DWDM) DCI system normally generates a large number of alarms when a fault occurs, but 95% of them are correlative alarms that result in numerous invalid work orders. With the AI-based aggregation of alarms functionality on Huawei OptiXtrans DC908 and NCE, root cause analysis can be quickly carried out.
Taken together, the updated set of features and capabilities designed into Huawei OptiXtrans DC908 make it a very strong leader in the DCI market. This product has been widely used in leading global Over The Top (OTT) providers and cloud service providers, and it has been widely deployed in other industries such as finance, power, energy, high-end manufacturing, and Internet Service Providers (ISPs).