A storehouse of solutions: options include speed, simplicity and cost savings. (Network Management).
Currently, storage capacity needs are doubling every six to 12 months. Business continuity/disaster recovery, enterprise resource planning (ERP), customer relationship management (CRM) and e-business efforts are requiring huge storage capacities, moving enterprises away from direct-attached storage, which does not allow data or capacity sharing between servers.
What are enterprises doing to solve their storage needs? Pacific Coast Jiffy Lube was looking for cost effectiveness, simplicity and speed. "The prospect of replacing our servers and workstations frequently made no business sense," offers Sean Porcher, its director of operations. "We wanted to implement a solution that could easily and cost-effectively scale along with our operation."
THE GROWING NEED FOR STORAGE
Enterprises are selecting data backup and storage solutions options that provide their users specific features for their particular networks. Pacific Coast Jiffy Lube chose network-attached storage (NAS), as did Aero-Metric in Sheboygan, WI. United American Insurance Co., McKinney, TX, upgraded its data backup and protection with a tape library. Systems & Methods Inc., Carrollton, GA, enhanced its system with backup and disaster recovery software. And pharmaceutical firm Apotex in Toronto settled on a storage area network (SAN).
At Pacific Coast Jiffy Lube, customer and inventory information stored at the corporate office is up dated nightly. Each service center must access this critical information 24x7 for day-to-day operations. A network infrastructure to store large amounts of continuously accessible data was needed.
"Adding servers would have been a very expensive alternative," according to Porcher. "Without a network administrator to manage our infrastructure, simplicity became very important. Last, we needed to improve the speed of our network, enabling faster data sharing between the corporate office and service centers. The qualities of network-attached storage fit ideally with our needs."
After a competitive analysis, the POPnetserver 2000 from the FIA Storage Systems Group, a San Clemente-based supplier of NAS systems, was chosen. "With a price of $895, the price-to-performance ratio was better than any other NAS system we evaluated," notes Porcher.
Overall storage capacity is 240 GB in the same 1U enclosure. The product was configured for the company with two 30-GB hard drives--and is expandable to a third--for a total capacity of 60 GB, and for RAID 1, or mirroring, which allows the second set of drives to duplicate the first set for maximum data protection.
"In 15 minutes and two mouse clicks, I had the system up and running," Porcher says. "The POPassist software, a Windows-based user interface for system administration, allows quick and easy configuration, and lets us remotely monitor and manage storage resources in real time." The POPnetserver is currently used across all Pacific Coast Jiffy Lube's departments, providing high-capacity storage for numerous workstations, and running the company's extensive inventory management system and accounting database.
"The speed and functionality of our entire local network has been improved," says Porcher. "Since the NAS system relocates storage onto its own independent platform and separates file sharing from application serving, file-server bandwidth was freed up and overhead on our existing application servers was reduced.
"The networks speed has increased drastically, saving up to two hours a day in network wait time. For a cost of approximately 1.5 cents per megabyte of network-attached storage, the system's value proposition is strong."
For more information from FIA Storage Systems: www.rsleads.com/208cn-258
Check out tape library
When United American Insurance Co. (UAIC) analyzed its storage needs, protecting data--its lifeblood-was an important consideration. New imaging systems, database stores of mainframe information, Web-based Internet and intranet information, and company growth were taxing existing data protection methods. Losing data meant losing time and money.
UAIC, a Torchmark Corp. subsidiary, specializes in life and supplemental health insurance products in 49 states, the District of Columbia and Canada through a large network of independent, as well as 2,700 exclusive, agents and 500 home office employees. UAIC stored files include policyholder information for 680,000 accounts, home and branch office business records, agents' support files, and traditional sales and marketing reports. Most company employees depend on the network. All corporate users phone in, e-mail with questions or access information through both Internet and intranet systems.
UAIC needed to fully prepare for quick recovery from any data disruption. The company's existing system, a single tape drive and two tape changers operating under a limited software program, backed up and stored the ever-growing volumes of data. As the plethora of data created and collected continued increasing, the extended time required to protect it was taking its toll on LAN performance and causing a rise in storage administration costs.
UAIC's large IT infrastructure--35 servers and more than 500 nodes--necessitated a partnership with StorNet Inc. The Colorado storage management and data protection solutions provider recommended an Overland Data Pro Tape Library--consisting of three modules with four AIT-2 drives and 57 tape slots--and Legato Networker software. This combination provided reliable, unattended backup functionality--thus saving staff time costs--and could be scaled to handle future growth. Each tape has a native capacity of 50 GB (100 GB compressed). When fully loaded, the unit can house up to 2.85 TB (native) or a total of 5.7 TB of compressed data.
With multiple tape drives, UAIC can restore files/servers faster and perform restores during scheduled backup operations. The backup and recovery infrastructure also integrates the tape library with UAIC's existing array of Windows 2000 and NT servers over the LAN's 100BaseT Ethernet interface. UAIC found that full backups, which once started on Friday evenings and ran into Monday mornings, are completed now in 24 hours. UAIC is realizing a 75% reduction in backup window time over the previous method.
For more information from SterNet: www.rslends.com/20Bcn-263
Backup pays off
Among the services Systems & Methods Inc. (SMI) provides is ensuring that intended recipients of child support payments receive money. Although state governments in Georgia, Connecticut, North Carolina and Missouri are responsible for collecting and forwarding that money to the appropriate recipients, SMI handles information processing for the state and local human services agencies. SMI maintains important databases--from beginning to end-using its own Systems and Methods Automated Remittance Technology, or SMART, solution.
In executing ongoing mission-critical jobs, SMI needed a dependable backup and disaster-recovery solution that could manipulate extremely large amounts of data on a daily basis, quickly and efficiently. "It's not uncommon for us to process 18,000 image files/transactions per site per day, so backups need to be fast." SMI also needed the capability to recover this data at any time.
"The reality of not having a reliable solution is more than just a scary thought for us," says Opie Thompson, network administrator at SMI. "The state governments we serve rely heavily on us, so losing data is unacceptable."
After experimenting with several well-known backup products, SMI eventually chose UltraBac's Enterprise Edition software as the backbone for its backup and disaster-recovery system.
SMI installed the software on approximately 50 servers, roughly 12 at each location. Now SMI administrators run nightly backups, often more than one at a time, at each site. Data from state case workers compiled using SMI's proprietary software, as well as SQL database files and Thompson says reliability is the solution's best feature, but fast backups and ease of operation run close seconds. The solution creates a log describing the tasks it performed, and is set to automatically e-mail Thompson to let him know the status of each server backup and whether any errors occurred during the process.
"Before, we had to really nurse our backup system, which meant working nights and weekends," says Thompson. "Now, I can automate our entire backup and disaster-recovery system and walk away, monitoring it periodically."
For more information from UltraBac Software: www.rsleads.com/208cn-262
SAN benefits pharmaceutical firm
Restoring files instantly in a competitive business environment, and eliminating all points of failure in the tape media backup process propelled Michael Davidson, chief information officer of Apotex, to propose a disk-to-disk (D2D) backup strategy for distributed servers. In the past quarter-century, the Canadian generic drug manufacturer and biotechnology firm grew from two employees to more than 4,000 employees occupying two million square feet in facilities in several cities. Its backup environment, however, looked like a server league of nations.
About 500 GB of data were backed up on 70 distributed servers in four operating systems--VMS, Novell, Windows NT and Unix, each with its own backup product--with 15 tape backup devices using either DLT or DAT media. The media library of thousands of tapes had to be maintained throughout various retention cycles. Eventually, the company settled on two server plat, forms--Windows NT/2000 and Sun Solaris. Storage more than doubled to 1.6 TB of data to back up.
Now, the Toronto data center has a Compaq ProLiant 530 server connected via Brocade switches to a Compaq EMA 1200 StorageWorks RAID storage subsystem, which scales to about eight terabytes. An identical secondary SAN in an affiliate's data center 30 miles away acts as an alternate. Another affiliate without a SAN is involved in the backup. A 100-Mbps fiber-optic data link connects these three sites.
D2D backup Agent software from EVault was installed on each SAN server to be backed up to the primary and secondary SANs, as well as on the workstations to be used for initiating backup-and-restore functions, and for receiving messages about the backup tasks. Backups automatically run in parallel during the night at differently scheduled times.
"This procedure enables us to have two SANs with identical copies of backup data online at all times," Davidson offers.
"We can easily retrieve it, and we also have a copy of the data stored off site for disaster-recovery purposes."
EVault's delta technology copies just the changes, not the entire file, since the last backup for that server. Each time a backup occurs, the software looks for files that have block changes, and then locates those changed blocks, compresses and encrypts them, and sends them over the wire. "A 50-GB file now takes about 20 minutes to backup rather than several hours using tape," Davidson says.
"Within a matter of minutes--not days--we turn around requests for restores," he says. "We realize that we'll need access to historical files stored on tapes off site. Our plans call for converting those tapes to disk media on the SANs. We'll use the existing backup software to read the files back to disk, and then the Agent software to back up these files to the SANs."
For more information from EVault: www.rsleads.com/208cn-256
Aero-Metric personnel operate 30 workstations digitizing and scanning aerial photographs, editing maps and images, and quality checking all deliveries. The company's storage-intensive services include color, black and white, and infrared aerial photography, along with photo scanning, image rectification, 2D and 3D digital mapping, geodetic and airborne global positioning system and inertial measuring surveys, LIDAR, digital elevation models and geographic information systems.
Aero-Metric completed a major network storage system upgrade by purchasing Procom's NetFORCE 3100HA network-attached storage filer with dual Fibre Channel data paths and active RAID controllers to centralize its disk storage needs. The filer is configured with a dozen 73-GB hard drives, and can scale as the company's data storage requirements grow--up to 60 drives, with a maximum raw capacity of 10 TB.
The NAS is used to store image files--some more than 3 GB--that are scanned from aerial photography. Daily, weekly, and monthly backups are performed with a Compaq Proliant Windows 2000 server running ARCServe 2000, with an external Quantum/ATL Powerstore M1500 20-tape DLT 7000 library.
A 3Com 4900 Giga-switch connects the NAS through fiber and the workstations through copper Gigabit Ethernet. The 3Com switch attaches to stacked Baynetworks 450 10/ 100/1000 switches connecting the remainder of the LAN.
"Information once dispersed over a number of locations is now aggregated into a central, fault-tolerant repository, giving users fast, easy access to needed files," according to Aero-Metric IT manager Michael Mertens.
The combination of the NetFORCE and Gigabit Ethernet increased response times dramatically. With single copy data sharing, the filer has simplified storage consolidation, and asset management, while streamlining configuration and reducing complexity. This results in further performance boosts, since fewer data translations and drive mappings are required.
"To make the NetFORCE generically usable took about 30 minutes," explains Mertens. "To put it into use for actual production took several days.
Ultimately, Aero-Metric has decreased storage costs further and now manages all of its data with minimum administration and maintenance.
For more information from Procom: www.rsleads.com/208cn-257
RELATED ARTICLE: Why choose NAS?
by Genevieve Ortegon
NAS systems are completely dedicated to storage, making them a viable solution for improving the speed and functionality of a network. NAS relocates the storage onto its own independent platform, separating file sharing from application serving. Since applications and storage are no longer running on the same system, this frees up file server bandwidth and reduces overhead on existing application servers.
Almost all NAS systems incorporate a feature-redundant arrays of independent disks-called RAID. RAID capability can protect and provide immediate access to data, despite a single or concurrent disk failure. Different levels of RAID offer different levels of protection. With RAID O, data is striped across all physical drives to improve access times; RAID 5 distributes data and parity across all drives and is capable of tolerating the loss of one drive, providing full drive integrity.
Many business infrastructures may contain a mix of Windows, Windows NT, Apple Macintosh, Novell Netware, Unix and Linux platforms. Traditionally, sharing data across these different platforms can be both challenging and expensive. With a NAS system, cross-platform sharing becomes simpler.
On the network, a NAS system can appear like a native file server to each of its different clients--meaning that files are saved on, as well as retrieved from, the NAS system in their native file formats. Converting the entire office to one single platform is not necessary, nor is losing the initial investment in desktops, servers and workstations.
Today's NAS systems are out of the box, plug and play. Installation does not require a high level of technical skill or a background in computer science. Users usually can create networked storage within minutes. Intuitive software programs guide in managing the network and getting the most out of the NAS system.
NAS is also an attractive option due to its cost. For a few thousand dollars, today's NAS systems offer the same performance, reliability and feature sets for which enterprises pay $10,000 or more for other solutions.
Ortegon is with FIA Storage Systems Group, San Clemente, CA.
RELATED ARTICLE: Manage the storage network.
by Robert Wright
By promising improved scalability, evailobility and data protection, SANs have emerged as an important data storage solution. As storage expansion gains momentum and dm storage infrastructure evolves, the challenge becomes storage management.
The initial, successful SAN implementation, based on one switch, is easily managed by the vendor-supplied application. As the SAN is expanded, a variety of vendors' devices is introduced, each bringing another management application, and each supporting different emerging standards and features. This increasing complexity and resultant drain un management resources threatens the primary promise of SANs--decreased storage-management costs.
A single, centralized storage resource management (SRM) application is required to perform the most significant tasks of managing the complete storage enterprise. Key SRM tasks include: discovery and visualization, device management, and performance and status monitoring that can be launched anytime, anywhere.
Visualization begins with automatic discovery of storage devices, switches, hubs, routers and host-bus adapters, including devices located on local and remote subnets. The SRM application recognizes each device, how it should he managed and monitored, and then assists in its integration into the storage enterprise. Information gained from the discovery process is saved so that changes may be recognized--a feature labeled persistence.
Using discovery process Information, a visual map of the storage enterprise is prepared, presenting all devices and their interconnections, and highlighting redundancies within the SAN, so management can visually confirm the required high-availability design, The SRM application simplifies the individual device management--either directly configuring each device, or, where necessary, automatically launching the appropriate device-specific application.
Comprehensive SRM applications continuously monitor every device in the storage enterprise. All events are logged and addressed as predefined, which may include paging IT management. Events range from minor warnings, such as a cooling fan RPM drop or a redundant power supply failure, to major events, such as a severed cable or complete device failure.
The SRM application will further regularly poll each device, ensuring that the loss of a device or connection is promptly reported. Logs provide the critical information for diagnosing problems or performance degradations. Performance and capacity information may also be tracked. User-specified thresholds allow SRM applications to notify users when certain conditions exist. This application should also collect reporting datato facilitate dynamic trend analysis, which helps optimize performance, and forecast capacity and consumption requirements, permitting management to proactively address storage demands.
For more information from McData: www.rsleads.com/208cn-255
Wright is vice president and general manager of software, McDATA Corp., Broomfield, CO.
Contributing to this article were Ron Levine, a computer information systems instructor at California's Santa Barbara City College, and Elizabeth M. Ferrarini, a freelance writer in Boston.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Pacific Coast Jiffy Lube|
|Comment:||A storehouse of solutions: options include speed, simplicity and cost savings. (Network Management).(Pacific Coast Jiffy Lube)|
|Author:||Levine, Ron; Ferrarini, Elizabeth M.|
|Date:||Aug 1, 2002|
|Previous Article:||The elevation of an IT outsourcing partnership: with HP SAP-trained staff, KONE's global strategy is top level. (Cover Story).|
|Next Article:||Conferencing: boardrooms to classrooms: the laptop-and-lunchbox generation learns that knowledge is the future's new commodity. (Telephony/CTI).|