Cloud computing Peaks Storage capacity

Will the rise of cloud computing subvert existing networked storage architectures? No one can tell now when this change will happen. To be sure, cloud computing presents unprecedented challenges to storage capacity, storage access performance, and cost.

The advent of cloud computing has facilitated the development of cluster storage technology

What is the important impact of the rise of cloud computing on information storage? "The starting point for cloud storage should be petabyte," says Ningyuxiang, EMC's chief architect at China Research and Development Center. ”

The rise of cluster storage

IDC research shows that from 2006 to 2010, the total amount of global information will increase by more than 6 times times, from 161EB to 988EB (1EB=1024PB). 750GB or 1TB SATA drives have been widely used in some of the newly launched disk arrays. The disk array, which is currently known to have the highest storage density, can provide up to 42TB of storage capacity in 4U space, which was previously unthinkable.

The latest generation of LTO-4 tape has a single cartridge storage capacity of 1.6TB (compression ratio of 2:1). The continuous progress of technology will promote storage to higher capacity development, and duplication of data deletion, compression and other technologies, can further enhance the utilization of storage space. In terms of performance, FC disk arrays have gradually transitioned to the 4Gb era, while 8Gb FC is waving to data center users; Gigabit IP storage is no longer an armchair; in the InfiniBand field, there are already manufacturers to launch 40GB InfiniBand adapter products.

Can existing networked storage architectures, such as Sans or NAS, effectively support ubiquitous cloud computing environments? Doubts were expressed. The main argument is that in the face of petabytes of massive storage requirements, traditional San or NAS in the capacity and performance of the expansion of the bottleneck; cloud computing, a new service model, inevitably requires a very low cost of storage architectures, and some of the existing high-end storage devices are clearly not enough to meet this demand. From Google's practice, they did not use the San architecture in the existing cloud computing environment, but instead used the extensible distributed FileSystem Google File system (GFS). This is an efficient cluster storage technology.

In recent years, the emerging cluster storage technology, not only easily break through the San performance bottlenecks, but also to achieve a linear expansion of performance and capacity, which is the pursuit of high-performance, high availability of enterprise users is a new choice. With the rapid development of some enterprises focusing on cluster storage business, such as Panasas, Isilon, long storage technology and so on, the application of cluster storage technology will become more and more popular. While cluster storage has an obvious advantage in dealing with unstructured data, it is unlikely that cluster storage will completely replace traditional networked storage in a short period of time, and that San and NAS can still be useful.

It is important to emphasize that virtualization is a core technology for achieving cloud computing vision, because cloud computing itself is a new generation of computing platforms that provide virtualization and high availability. From the current market situation, server virtualization is already in full swing, while the development of storage virtualization is relatively slow. At the end of 2007, EMC launched the SAN storage virtualization Product Invista 2.0. Invista 2.0 supports 5 times times more storage capacity than previous generations, further increasing availability, enhancing data protection and management capabilities, increasing efficiency, and enhancing scalability. In addition, Invista 2.0 also has VMware certification, which allows users to better manage, share, and protect information in the VMware architecture.

"To effectively support cloud computing, the infrastructure must have several key features," said Craig Nunes, a storage company's vice president for 3PAR marketing. First, these systems must be autonomous, that is, they must embed automated technology, eliminate manual deployment and management, and allow the system to respond intelligently to application requirements. If the system needs human intervention to allocate and manage resources, it cannot adequately meet the requirements of cloud computing. Second, the cloud computing architecture must be agile and able to respond to demand signals or changing workloads in a timely manner. In other words, embedded virtualization technologies and clustering technologies must be able to respond to rapid changes in business growth or service level requirements. If the system takes hours, days, or weeks to respond to new applications or user needs, the system will not be able to meet the requirements of cloud computing. ”

SaaS is also the abbreviation for storage as a service, meaning storage. In the cloud computing environment, storage is no longer a cold hardware device, but a service. Will this change the way in which future users store their purchases, from purchasing hardware to buying storage services? "Within a large enterprise, whether it's using cloud computing or building a common data center, end users ' It requirements will eventually be serviced," says Craig Nunes. ”

EMC bullish on cloud storage

A message that has been rife on the Internet over the previous period is that EMC and SAP will work together to refine EMCs SaaS strategy and adopt a host service called EMC Cloud. Ningyuxiang declined to comment because this was not an official EMC announcement. However, it is true that EMC is moving into the cloud computing world.

February 2008, EMC announced the acquisition of Pi for cloud computing vendors. This is a manufacturer specialized in providing software and services related to personal information management. PI's products and services encompass a range of cloud computing features, such as distributed replication of user data. After a recent series of acquisitions, EMC adjusted its internal organizational structure to 6 of the original 4 (including storage, content management, VMware, and RSA), and the new two departments are the consumer/Small Business product division and the Cloud Infrastructure and services Division. According to Ningyuxiang, at present only the EMC China Research and Development Center has more than 100 technicians specializing in cloud computing research and development work.

Online storage services are a typical application of cloud computing. According to IDC, the global market for online storage services will reach $715 million trillion in 2011. Also an EMC-acquired company, Mozy, which has been offering online backup services two years ago, now has hundreds of thousands of of paid customers, not only for individual users, but also for large business customers like GE. Fan, global vice President of EMC Inc. and general manager of China Research and Development Center, is a loyal user of mozy.

Ningyuxiang stressed: "The key to the popularity of cloud computing is the user experience." For end users, they don't care what kind of technology is used backstage, they can get the desired result through simple operation is their real concern. Maybe one day, no matter where you go, just carry a small thing like u disk, which contains personal security authentication information, as long as you plug it into any terminal equipment, you can get the information you want from the ' cloud '. ”

Face the problem of standardization

When it comes to the concept of cloud computing, many people naturally associate the concepts of grid computing and utility computing, which seem to have some shadow of cloud computing. But there are professionals who think that cloud computing is different from utility computing and Grid computing, the rationale is that grid computing usually refers to a resource pool environment (such as image processing) for computing tasks, and utility computing is a resource pool environment for mastering long-running processes and tends to focus on meeting service levels While cloud computing in most cases refers to the different kinds of services available on the network, and cloud computing can be run in Grid computing or utility computing environments.

"From Grid computing, utility computing to cloud computing, walking is a spiral of development." Cloud computing has made a clear-cut and emphatic emphasis on the importance of virtualization technology, which was not previously. "The concept of cloud computing is really very vivid," Ningyuxiang explains. The so-called cloud is in the air, which means that the relevant computing or storage work is not done on the local device, but in the air or at the remote end. ”

Understanding cloud computing can capture two key elements: first, cloud computing is based on distributed computing; second, cloud computing provides a new business model in which Third-party service providers with the advantage of talent, money, services and the ability to scale can emerge in large numbers. "The essential feature of traditional computing is that computing and storage are done locally," Ningyuxiang. Enterprises with traditional computing methods will permanently retain the relevant software and hardware. The cloud computing approach is not, it is by third-party providers to provide computing and storage resources, and is responsible for the operation and maintenance, users only through a certain standard of terminal tools (such as browser) access to the system, you can get the required services. ”

At present, the industry's research on cloud computing is still in its infancy, and the rapid development of cloud computing relies on the development of standardization to some extent, otherwise, there may be communication obstacles between the "cloud" and "cloud" of different manufacturers. Although the application of cloud computing is still lack of comprehensive technical standards, but we also see some manufacturers have made a useful attempt in this regard. 3PAR, for example, has launched a development program called 3cV, which combines 3PAR Inserv storage servers, HP C-class blades and VMware architectures to form a complete public computing product that helps users reduce server, storage, and operational costs by half At the same time getting better agility, Deutsche Bank has successfully implemented the program.