Enterprises usually build multiple information centers to ensure Disaster Recovery (DR). The active data heart transmits users’ core services. Backup data centers transmit some not-primal services, and backup data, configurations, and services of the active information eye. The active data center and the fill-in data centers piece of work independently. If a disaster occurs and the agile data center cannot provide its core services, then the backup data centers tin can quickly recover data and applications. This reduces any losses caused by the disaster.
A Data Center Interconnect (DCI) network plays a disquisitional role past transmitting services with one of three interconnection methods: Layer i optical transmission interconnection, Layer two interconnection, and Layer 3 IP interconnection. Different services accept different interconnection requirements. Hierarchical interconnection and disaster recovery create data centers that feature High Availability (HA) interconnection and service continuity. In add-on, server clusters can be installed in unlike information centers to enable disaster recovery for application systems.
Data-level and service-level disaster recovery take different requirements. For example, storage networks require low latency, large bandwidth, and loftier reliability, and service networks require link backup and fast route convergence.
Layer 1 interconnection:
Features low latency, big bandwidth, high security, and reliability to provide real-time data centre disaster recovery
The rapid development of deject calculating and Big Data requires that enterprises provide ‘uninterrupted services’ and ‘no information loss.’ To exercise this, data centers
must provide real-time backup mode to ensure data synchronization between the storage systems of active and backup information centers. This ensures that customers will non experience any latency in service processing speed and that data is not lost.
Layer 2 interconnection:
For HA clusters in different sites, the states of cluster nodes are maintained and controlled through heartbeat signaling and data synchronization.
Virtualization technology is becoming increasingly important for disaster prevention. Virtual Motorcar (VM) migration plays an important role in the expansion of information heart capacity. Information center services transmitted by VMs must non be interrupted during VM migration. Land synchronization and maintenance sessions require that physical servers be located in the same subnet prior to and subsequently VM migration. Therefore, cross-site VM migration
also requires Layer 2 network interconnection.
Still, at that place still are problems that must be solved; Poor adaptability does not allow Ethernet technology expansion based on different networks; just a limited number of information eye interconnections are possible; Flooding of unknown IP address traffic and suboptimal paths causes a considerable waste of WAN link bandwidth; Multi-homing admission is non available. Traffic forwarding does non enable load balancing; Network configurations are tedious, and Operations and Maintenance (O&M) is complex.
Layer three interconnection:
Preventing disruption of services caused past a single information centre disaster has go the focus of enterprise users. Data centers require high levels of protection against risks that include loftier volumes of information aggregation and natural disasters such equally storms and earthquakes. Data center disaster recovery is divided into data heart, service-, and application-level disaster recovery. Specific disaster recovery mechanisms can exist selected based on users’ investment and service growth demands.
HuaweiCloudFabric Solution (CloudFabric) provides highly efficient, reliable iii-layer interconnection among multiple information centers that satisfies the interconnection and disaster recovery requirements of a wide multifariousness of information center services.
Ethernet Virtual Network: Highly Scalable and Efficient Layer 2 Interconnection
The Huawei CloudFabric EVN’s industry-leading Layer ii DCI solution provides highly permits scalable, efficient layer 2 interconnection. It permits expansion up to 32 data centers across IP WANs. EVN integrates multiple data centers into a super-big IT resource pool that not but accommodates inter-data center server clusters, only as well permits free migration of 256,000 VMs beyond data centers. Huawei’south EVN solution features high availability, scalability, and bandwidth efficiency, too equally simplified O&M. It tin can completely address Layer 2 interconnection issues and fully optimize control, data, and management planes.
Uses the Border Gateway Protocol (BGP) and supports interconnection amongst up to 32 data centers; Optimization of unknown IP accost traffic prevents inefficient WAN link bandwidth; Provides 256 m MAC addresses that accommodate the migration of over one million VMs
Uses the VXLAN protocol, which permits expansion to 16 tenants; The unabridged network supports traffic-based load balancing for optimal efficiency
Requires but four configuration commands to simplify network deployment; Provides automated and visualized network deployment
Huawei CloudFabric MPLS VPN provides highly reliable, secure Layer 3 interconnections. Information technology enables a large number of users to gain admission to data centers and permits Layer iii services access among data centers. Other features include:
End-to-stop 200 ms protection switching and 99.999 percentage reliability
Transmits multi-tenant services in a unified style and isolates services of different security levels amongst different tenants or of a single tenant
Provides five-level Hierarchical Quality of Service (HQoS) scheduling. Unlike tenants and different types of services of a single tenant can obtain unlike levels of Service Level Agreement (SLA) guarantees. This satisfies the requirements of users for loftier bandwidth and low latency, jitter, and packet loss ratio.
Fully supports Option A (dorsum-to-back), Choice B (single-hop and multi-protocol), and Choice C (multi-hop and multi-protocol) for any networking scenario
Optical Transport Network (OTN) Interconnection
Huawei CloudFabric uses a large-capacity, depression-latency optical transmission interconnection solution to provide Fibre Channel Storage Area Network (FC SAN) and Internet Protocol Storage Area Network (IP SAN) arrangement interconnection and disaster recovery.
Huawei Optical Switch Node (OSN) serial OTN devices provide manufacture-leading WAN transmission capabilities. These devices tin form loftier-speed, existent-time Wavelength Partition Multiplexing (WDM) optical networks. They also tin can exist used by disaster recovery systems that require large capacity and low manual latency, and can transmit massive amounts of data. For example, a single cobweb provides upwardly to 8 Tbit/s bandwidth. OSN series devices support xiv types of SAN interfaces (such as FC, FICON, and ESCON) and have earned compatibility certificates from seven mainstream storage device vendors. OSN series devices can establish a SAN that enables the longest disaster recovery distance in the industry (3,000 km). They also provide carrier-grade (50 ms) protection switching.
‘IP + Optical’ Disaster Recovery
‘IP + Optical’ disaster recovery collaboration: Fast route convergence and quick fault recovery (less than 50 ms) heighten data- and service-level disaster recovery capabilities. They likewise accommodate requirements for data eye disaster recovery in the same or different cities:
Disquisitional DR Indicators
The 2 almost critical indicators for measuring disaster recovery are Recovery Time Objective (RTO) and Recovery Signal Objective (RPO). The ultimate goal of all disaster recovery solutions is for the value of these ii indicators to exist as close to zero as possible.
RTO indicates the time from when services are interrupted due to an IT organization breakdown to the time when the IT arrangement recovers and can support operations of all departments. RPO indicates a recovery indicate at which both system and awarding data can be recovered to support service operations of the entire department. The extent to which the system and application data are recovered can be the final week’southward backup information or real-fourth dimension data of the latest transaction.
Proposals for Levels 1 to 4 transportation-/information-level backup.
Proposals for Level four disaster/recovery backup.
Proposals for application/data-level disaster recovery deployment for Levels 5 and vi.
Proposals for distributed deployment of Level 6 awarding-level disaster recovery services.
A data middle — an integrated ICT application environment resulting from the concentration of data — combines computing, network transmission, and storage resources, and has become the most critical infrastructure for an enterprise’south concern operations. The goal of enterprise IT personnel is to ensure service continuity and security of information and services, maintain competitiveness, and improve users’ experience. Huawei, a leading global network solutions provider, is dedicated to researching and developing the industry’s best data center interconnect and disaster recovery solutions.