Communicate
Solution--Making SAN Transport Network More Robust
Issue 24 (Topic on Optical)

By Li Hanguo

Only Professionals Can Put Customers at Ease

Information is a strategic resource for modern enterprises and is even a matter of life and death for some enterprises that have high requirements in terms of reliability and security. Hence, it should come as no surprise to find that many of these enterprises are willing to invest heavily in the construction of a disaster restoration system.

Ms Yang, the CIO of one of the more prominent financial institutions in Shanghai, is quite concerned about the normal operations of a disaster restoration system. Yet in recent months, she has been annoyed by problems that have frequently occurred in the SAN (Storage Area Network) of her own enterprise, and even more so, by the slow response from technical support engineers.

It was by happenstance that Ms Yang learned about Huawei's SAN transport solution based on Metro WDM. She subsequently invited a number of experts to conduct a comprehensive study and evaluation of Huawei's solution. At the same time, Ms Yang also carried out a bit of in-depth research on Huawei's after-sales service system. After careful comparisons, Ms Yang finally selected Huawei's OptiX Metro6100 SAN transport solution for the new SAN of her enterprise.

Actually, what satisfies Ms Yang most about the Metro6100 is the carrier-class, or higher reliability that is guaranteed, by its multi-layer redundancy architecture, which means that services can still run normally, even if faults occur at more than one point.

In addition, the extreme care and effort in the work carried out by Huawei's engineers greatly impressed Ms Yang. For instance, they performed further maintenance on fiber jumpers after troubleshooting multiple hidden problems in the optical fibers. So, no matter how unfavorable the weather or traffic condition may be, Huawei's technical support engineers always respond within several minutes after being contacted! Hence, Ms Yang has come to the conclusion that only true professionals can put customers at ease.

This is just one of many cases where customers have indicated satisfaction with Huawei's products and after sales service. Thus, more and more customers across the globe are now enjoying the services provided by Metro6100, and the dedicated technical support that comes with it.


Which SAN Transport Solution should Customers Choose?

The dedicated SAN network has become a must for information-oriented modern enterprises. As a matter of fact, the total volume of Metro WDM networks intended for SAN is rapidly increasing. SAN transport networks basically fall into two categories: dedicated networks and bandwidth leased networks. These two types of networks are able to cater to requirements for data storage with different reliability classes.

Placing importance on high reliability, large enterprises such as, financial institutions, generally adopt dedicated optical fiber networks. In contrast, small and medium-sized enterprises, or local offices, largely favor leased services, which are provided by telecom operators, for data centralization and remote data backup.

  • "Dedicated networks" to meet high performance requirements

    For dedicated networks, customers usually require a SAN with high reliability (class 5-7). Their main concern is the overall performance indexes of the networks, which include such aspects as, integrity, reliability, scalability, capacity of development, management and maintenance, and being able to strike a balance between various factors in the entire life cycle. Some of these indexes are even higher than those for carrier-class services. In short, the requirements for SANs are as follows: high stability, scalability, high QoS, and reasonable configuration.

    How does one go about designing a transport solution for such a high-performance SAN? The SAN referred to in Ms Yang's case, is a typical example. Since the SAN functions as the core enterprise in the financial industry, requirements and demands for higher performance are far greater.

    The performance requirements for the SAN include: up to 99.999% availability, a fault restoration duration of less than half an hour, no single-point fault, sustainable scalability, and support of new services emerging in the next five years.

    In view of these requirements, five key factors need to be taken into account: the grouping of services, the planning of NEs (Network Element), the choosing of protection schemes, the bearing of services, and the budgeting for optical fiber links.

    Why is it necessary to divide services into groups? In a high-performance SAN, services are basically configured as dual-homed or hot standby services. Only by dividing services into two groups can they be transmitted, using separate routes. In this way, the advantages of dual-homed and hot standby protections are fully utilized.

    How should enterprises conduct NE planning? In the SAN, two sets of equipment are required for one node (1+1 backup of the equipment). The reliability of the network is higher if the dual-homed route is independent of the hot standby route. However, if the dual-homed and the hot standby services are transmitted using the same equipment, then the dual-homed and hot standby protection capability of the routes is limited. Consequently, network reliability is reduced. In addition, 1+1 networking can be developed into two-plane network solutions, which can not only raise network availability, but also effectively expand the wavelength, thus, satisfying the requirements for development of future services.

    What so kind of protection solutions should enterprises choose? First, a comparison of the three main protection modes needs to be done (refer to Table 1). When considering the high availability requirement for the SAN, enterprises can choose inter-board optical sub-network connection protection. Although the cost of such a protection mode tends to be rather high, when it is combined with the 1+1 equipment protection, the system can ensure that no single-point fault occurs in the system, the board, or the modules. As a result, network reliability is much higher than carrier class reliability.

Table 1 Comparison of protection modes

Protection modes
Number of boards
Optical fiber links
Number of optical multiplexer/
demultiplexer
Reliability
Switching time
Cost
OLP protection
Smaller
Cold standby, unable to ensure that the standby optical fiber is available to high-rate services.
One pair in one direction
Average: The optical fiber links can be protected but many single-point faults may occur.
50ms
Very low
Optical sub-network connection protection (board embedded)
Smallest
Hot standby, able to ensure that the standby optical fiber and the receiving optical module are in the normal state.
One pair in either direction
Good: The wavelength and the receiving optical module can be protected, but the transmitting optical module and boards cannot be protected.
50ms
Very low
Optical sub-network connection protection (inter-board)
Largest
Hot standby, able to ensure that the standby optical fiber and the boards are in the normal state.
One pair in either direction
Best: All points can be protected and no single-point fault occurs.
50ms
High

How does one design a budget for optical fiber links? To accommodate new services in the future, the overall SAN transport system is designed to support 10G services. Considering hidden troubles detected on the optical fibers during the engineering survey and also the inconvenience in maintaining the leased optical fibers, Huawei recently introduced its optical power control technology over the lines. The system can continue to work normally, even if optical fiber loss is greatly increased, by using this technology.

The entire SAN system is designed on the basis of Metro6100. The system features a reasonable configuration, clearly-defined layers, high redundancy, stable running, and good scalability. Furthermore, the SAN system, in combination with Huawei's professional services, can completely satisfy the data storage and transport requirements of high-end financial institutions.

  • "Bandwidth leasing" to meet low-cost requirements

    Due to limited SAN bandwidth, small and medium-sized enterprises usually choose to lease bandwidths, instead of constructing their own dedicated SANs. In this way, they are not only able to reduce the cost of construction, but also the effort required to maintain dedicated SANs.

    Large enterprises that need to transmit storage services over several hundred kilometers are also likely to lease bandwidths, as construction of their own dedicated networks often demands heavy investment. Moreover, telecom operators are also required to provide a new type of service, which is storage backup, and they now include bandwidth leasing as an integral part of their service packages.

    These storage services generally demand carrier-class reliability. Therefore, transport at the carrier-class must be guaranteed. Considering the return-on-investment, operators usually hope that a single wavelength will be able to provide the highest bandwidth. Thus, they must strike a balance between carrier-class reliability and higher bandwidth.

    To help solve this problem, Huawei recently developed a 10G board for Metro6100. The board is capable of supporting a maximum of eight FC and GE mixed services and will be able to further support the latest FC400/1200 services and remote module in the near future. As a result, one or two wavelengths can satisfy the needs of a small or medium-sized enterprise for local data storage and transport, as well as the needs of a large enterprise for remote data storage and transport, and even the needs of a telecom operator for data storage backup services.

    For operators, one or two types of OTU boards, combined with a tunable feature, can swiftly and easily respond to such market demands, thus, allowing the TCO (Total Cost of Ownership) to be effectively controlled. As for customers, the huge and unaffordable construction expenditure can be reduced. In addition, reliability can be guaranteed and heavy maintenance costs can be avoided, thus, truly making this a win-win solution for everyone.


Metro6100 Makes SAN Transport Network More Robust

Applicable to different network layers, ranging from the trunk layer, to the access layer, Metro6100 is currently being widely employed by some of the world's top operators such as, BT and KPN, and has already passed all the major certification requirements. Providing hardware and software verified by mass traffic tests and pooling Huawei's excellent service, Metro6100 allows customers to comfortably exchange their views with Huawei's experienced network experts, which in turn, allows them to obtain more robust and customized SAN transport solutions.




Copyright © Huawei Technologies Co., Ltd. 1998-2009. All rights reserved.