Store those Things (iv) traditional commercial storage into the cloud computing road

Started learning Hadoop a year ago. Hadoop, the fact standard for large data, is widely used by Facebook, Yahoo,ebay,taobao,baidu. Hadoop itself is complex and contains many subprojects, and different scenarios spawn different subprojects.

Recently to do OpenStack project, began to learn OpenStack. Also reflect on what I am now engaged in the industry: Enterprise Storage, where the road?

It's technology development can be said to be changing. Google abandoned the MapReduce architecture a few years ago, when it was not yet gradually perfecting Hadoop. Now Google's new troika caffeine, Pregel, and Dremel are said to be able to control 1PB query requests within 3 seconds. This technology I have not been to understand. But to be sure, the impact of the new troika on it is no more than the three of Google papers MapReduce, BigTable and gfs a few years ago.

The technological development of traditional commercial storage can be said to be quite slow. Many of the proposed models are based on business rather than technology. For cloud computing and large data, traditional commercial storage is not a good way to go, but to integrate, both to be able to store their own storage as the data center of the cloud computing age. This article also explores how traditional commercial storage can be supported by cloud computing. Here we will refine the cloud, just to discuss OpenStack and Hadoop. Of course this is a bit one-sided. But since no one can give the academic definition of cloud computing, our one-sided is not a big mistake.

Because I don't know much about other friends, I'm just here to discuss how EMC's VIRP integrates EMC's storage arrays into Haoop and OpenStack.

VIPR is an EMC software-defined storage product VIPR, expressed as virtualization Platform reinvented. VIPR is not a physical storage system, but a logical storage system. VIPR combines EMC and other third-party storage into one storage pool, managing the storage pool as a single system while preserving the full value of the underlying specific storage. VIPR can span multiple locations and data centers, use multiple different storage platform offerings, provide unified block storage, file storage, and object storage services, while VIPR provides a unified metadata service and self-service deployment, metering, and monitoring services. In addition, VIPR is also available for multi-tenant environments.

VIPR employs an open Scale-out architecture that is deployed on a 3-node cluster that delivers high availability, load balancing, and system-free upgrades for large, complex storage environments. The VIPR interface has RESTFULAPI, GUI (Console), CLI, and SDK to provide maximum flexibility to users.

VIPR can map heterogeneous physical storage System abstraction in the datacenter to a large logical shared storage pool, providing great value to enterprise IT departments and cloud service providers. This is a great improvement for storage resource management and mixed cloud deployments in geographically distributed data centers. VIPR can effectively and flexibly manage the data-centric cloud infrastructure.

VIPR addresses one of the big issues facing current enterprise IT and cloud administrators: How to move multi-vendor storage systems to a cloud infrastructure and manage and share storage systems from a central view while still using and maximizing the capabilities of existing storage systems. VIPR simplifies storage operations in the data center and provides services for block, file, and object storage data through a self service management approach.

See more highlights of this column: http://www.bianceng.cnhttp://www.bianceng.cn/database/storage/

SOURCE anzhsoft:http://blog.csdn.net/anzhsoft/article/details/20541623