By: Christine Harrell
Today's computing networks typically have very powerful client computers with significant amounts of captive storage attached to each one. Processing power is very easily available and the Internet interconnects a huge number of private networks and individual computers. Captive storage for clients, servers and dedicated applications has exploded. Successfully executed backups and (more importantly) successful restores, may mean the difference between gracefully recovering from data loss and starting the countdown to seeking new employment.As processing power proliferates, and dedicated captive storage follows along with each new machine, storage "islands" are gradually formed. Islands are dedicated storage arrays or volumes that are only used by a single host or a single application. One example of a common island is an Exchange server with an external SCSI array; another could be a SQL server with some dedicated RAID-protected drives. Captive locally attached storage for a database cannot be shared out to other computers nor can it be easily grown, moved or duplicated without placing a significant burden on or even causing downtime for the host. These islands represent wasted hardware investment dollars, since many times some are underutilized while others are overflowing. Also these islands propagate over time they increasingly burden the systems administrators with management of the new growth. Storage islands often cause backup problems, requiring dedicated tape drives and software to manage multiple individual backup tasks - not to mention the time and burden on the host when a restore is necessary.The benefits of centralized storage are that application data can be combined into a single storage device or pool of devices called a Storage Area Network (SAN). SANs provide very high performance data storage that can be scaled as needed. File sharing services, usually proved by partially or wholly dedicated servers, can be consolidated into a Network Attached Storage (NAS) server. A single application server can share files as well as host dedicated application data, as well as run software for backup and restore tasks. But as servers grow in capacity while narrowing their specialization, the advantages of consolidating become clearer.With increasing pressures of security, regulatory compliance and corporate governance, management of business data is becoming ever more complex and important. Technology is allowing organizations to create and store exponentially more data; the key to being successful is managing that data explosion.Consolidation means reducing the amount of time spent managing tasks, jobs,applications and data growth. Moving all types of data to a single, central, redundant and high-performing storage server could mean less time spent handling repeated storage-related tasks and worrying about backup failures.Direct Attached Storage (No sharing of data resources)The most common form of server storage today is still Direct Attached Storage (DAS). The disks may be internal to the server or they may be in an array that is connected directly to the server. The storage may only be accessed through that server. An application server will have its own storage; the next application server will have its own storage; and the file and print servers will each have their own storage. Backups must either be performed on each individual server with a dedicated tape drive or across the LAN to a shared tape device consuming a significant amount of bandwidth.Storage can only be added by taking down the application server, adding physical disks and rebuilding the storage array. When a server is upgraded then its data needs to be migrated to the new server.SAN (Sharing of data resources)A SAN allows more than one application server to share storage. Data is stored at a block level and can therefore be accessed by an application, not directly by clients. The physical elements of the SAN (servers, switches, storage arrays etc.) are typically connected with Fibre-Channel. Backups can be performed centrally and can more easily be managed to avoid interrupting the applications. The time taken for backups is dramatically reduced because the backup is performed over the high-speed SAN and no backup traffic ever impacts users on the LAN. The primary advantage of a SAN is its scalability and flexibility. Storage may be added without disrupting the applications and different types of storage may be added to the pool. Adding storage capacity has become more simplified for systems administrators, so it is no longer necessary to bring down the application server. Additional storage can simply be added and then configured and made immediately available to those applications that need it. Upgrading the application server is also simplified; the data can remain on the disk arrays, the new server just needs to point to the appropriate data set.The actual implementation of a SAN can be quite daunting given the cost and complexity of Fibre-Channel infrastructure components. For this reason, SAN installations have primarily been confined to large organizations with dedicated storage management resources.The last few years have seen the emergence of iSCSI (which means SCSI over IP or Internet Protocol) as a new interconnect for a SAN. iSCSI is a lower cost alternative to Fibre-Channel SAN infrastructure and is an ideal solution for many small and medium sized businesses. Essentially all of the same capability of FC-SAN is provided, but the interconnect is Ethernet cable and the switches are Gigabit Ethernet, the same low-cost technology that is commonly in use today on most LANs, slightly lower performance but most organizations will not notice.Network Attached StorageA NAS appliance is a simplified form of file server; it is optimized for file sharing in an organization. Authorized clients can see folders and files on the NAS device just as they can on their local hard drive. NAS appliances are so called because they have all of the required software preloaded and they are easy to install and simple to use. Installation consists of rack mounting, connecting power and Ethernet, and configuring via a simple browser-based tool. Installation is typically achieved in less than half an hour. NAS devices are frequently used to consolidate file services. Toprevent the proliferation of file servers; a single NAS appliance can replace many regular file servers, simplifying management and reducing cost and workload for the systems administrator. NAS appliances are also multi-protocol, which means that they can share files among clients using Windows and UNIX-based operating systems. Administrators manage the NAS device via a browser window from anywhere they have network access and can assign shares, security settings etc. NAS fits right in with existing security and network management tools. As the business grows and needs more capacity, more storage can be added to the NAS device without disrupting users. One of the most common requests to a systems administrator is to restore a single file or group of files. With NAS, a feature called Snapshot™ is available that provides an almost instantaneous way for the systems administrator (or even an authorized user) to recover lost, deleted or corrupted files.
Article Directory: http://www.article-tree.com
สมัครสมาชิก:
ส่งความคิดเห็น (Atom)
ไม่มีความคิดเห็น:
แสดงความคิดเห็น