In 2005, reliability was the watchword

The year 2005 was a time in which SNW Online started out talking about reliability, followed with a veritable feast of cutting-edge technology and concluded with a universal call for more reliability.

Throughout, the focus was on the needs of storage networking users, the men and women who typically catch the blame when things go wrong but somehow miss the praise when things go right.

Reliability? Yes!

In an article posted Jan. 31, Dick Benton, a senior consultant with GlassHouse Technologies, went on a reliability rant inspired by the then-red-hot and emerging ATA drives.

"With the advent of hot-sparing, various RAID schemas and vendor software supporting automatic rebuilding," he postulated, "is disk reliability still important? The answer, in most cases, is a resounding, 'Yes.'"

Benton went on to discuss worst-case scenarios ("It could take hours, if not days, to completely rebuild a 250GB disk in the case of a drive failure that initiated a hot-spare movement into the production environment"); the impact of regulatory compliance ("Is data on a failed RAID 5 disk potentially retrievable and readable?"); and the long-term potential of down time ("If there are 1,000 drives, 1% failure means the contents of 10 drives must be rebuilt and the disposal of failed drives certified").

Nielsen Media Research: NAS and SAN unite

A "technology smorgasbord" would be a good way to assess our Feb. 21 case study describing how Nielsen Media Research reacted to the steady growth of the IT management of its Windows systems. Nielsen started out by planning a consolidated NAS storage environment, but that limited configuration morphed into a combined, multi-tier, multi-vendor NAS-SAN storage infrastructure that supports numerous corporate business units while providing a foundation for a variety of innovative storage strategies.

In an effort to avoid the creation of separate NAS and SAN storage pools, the company implemented a NAS gateway solution that enabled the NAS storage to take advantage of the SAN's disk capacity, says Nielsen technology strategist Robert Stevenson.

The bottom line, according to Stevenson: "By consolidating NAS and moving the NAS data through a highly scalable gateway to a SAN that contains multiple storage tiers, Nielsen Media is creating a storage infrastructure that will allow the company to both control costs and grow."

OSF Healthcare: Real, live ILM

When it comes to technology smorgasbords, our Feb. 28 case study showed that OSF Healthcare is no wallflower. To get his organization's storage budget under control, technical planning manager Jim Doedtman's group worked with EMC Corp. to design a four-tier storage architecture that could support an enterprisewide information lifecycle management strategy.

Noted Doedtman, "The idea is to have the right information on the right storage tier throughout its lifecycle."

Evidently, the idea worked, because 65% to 75% of all Microsoft Exchange files were being moved from Tier 1 Symmetrix to Tier 4 Centera, enabling the organization to realize major cost savings. In addition, virtually all PACS back-end archiving will be done on Centera.

"Putting everything on disk and using disk-to-disk replication eliminates the human factor in managing information," Doedtman said.

How to select a backup vendor: Ask these questions:

On March 21, Bud Stoddard, president and CEO of AmeriVault, gave eight suggestions for selecting a backup vendor. In a nutshell, he suggested that prospective customers do the following:

  • Find out how long the vendor has been in business.

  • Visit the company's data center.
  • Find out if the vendor has other customers like you.

  • Make sure the company has a proven 365 x 24 x 7 customer service team.

  • Ensure that backup is its core business.

  • Check out the company's financial stability.

  • Insist on a service level agreement.

  • Move on if the price looks too good to be true.

SATA: Here to stay

In our April 4 article entitled, "SATA is here to stay," Barbara Murphy, vice president of marketing at Applied Micro Circuits, wrote that doubts about SATA's viability were largely dispeled because RAID technology provided the added fault tolerance required by IT managers. Citing SATA's combination of affordability and performance, she argued that storage priced for the desktop but capable of enterprise-level performance and reliability "is far too attractive to resist" when used in appropriate applications.

Murphy said that SATA is optimal for data — including e-mail — that consumes a lot of storage but is not accessed constantly. She also said the technology is a good fit for specialized business or scientific applications that require both high capacity and fast writes. Examples include storing large CAD files, imaging, natural resource and grid computing. And even though SATA has been viewed as a replacement for tape, she said that RAID-enabled SATA systems can be complementary to tape.

Continuous data protection: Moving into the mainstream?

Continuous data protection is another topic that got hot in 2005, and in a May 23 piece, LiveVault president and CEO Bob Cramer explained that rather than backing up systems every night, CDP captures data changes as they occur, saving them in many point-in-time, restorable versions throughout the day. From its cost-prohibitive roots, Cramer said, CDP is finally moving from "nice-to-have, to mainstream for business."

He went on to note that the best CDP solutions completely automate and integrate continuous disk-based backup, off-site protection, archival and Web-based recovery processes, eliminating both the opportunity for human error and the risk of physical media error or loss. "The result," Cramer claimed, "is 100% recoverability of the point-in-time version you need vs. today's high tape-based backup failure rates."

Hiring SAN experts: Does it ever get easier?

What's a year of coverage without an article on the challenges of finding highly qualified technicians? In our May 15 exploration of this topic, we described the difficulty of getting and keeping the requisite skills required for storage organizations, especially as these skills apply to SAN environments. The situation is complicated by the heterogeneous nature of SANs and the many pieces of technology they encompass, including networks, operating systems, databases and storage components such as Fibre Channel switches.

So what's the solution? For many companies, it lies in the ability to train systems, network or Unix administrators. However, Bob Shin, a principal at State Street Bank in Boston, eschews this approach, choosing instead to simplify State Street's storage environment by reducing the number of operating systems. Even more important, he obtains the necessary in-depth technology know-how by leveraging support from his vendors. This, Shin says, frees him up to hire people with an even more crucial skill set, which includes the ability to align storage capabilities with business needs.

Automated provisioning: SRM to the rescue

Automated provisioning: Depending on who you talk to, it's so easy to do that it's yesterday's technology topic. Not so, says Ed Palmer, director of product management at Storability Software, in his May 16 article entitled, "SRM steps in where automated provisioning has yet to tread."

According to Palmer, the current crop of automated provisioning products represents a step in the right direction but are "incomplete, restrictive and often inflexible," performing only a small portion of the complex provisioning process and requiring administrators to relinquish too much control over their infrastructures.

Enter SRM software. Among its many qualities, it can do the following:

  • Create actionable e-mail notifications to systems administrators when key resources and applications are running out of space.
  • Instantly identify which applications need storage and where available resources currently exist, providing a single, consolidated view of an entire multi-vendor, multi-site storage infrastructure.
  • Walk users through the process of finding available storage and disk volumes and even generate a provisioning order.

To recap, Palmer declares, SRM software provides the information needed to quickly make the right provisioning decisions and to take appropriate action in a timely fashion — while maintaining desired control.

That takes us halfway through a year in which reliability ruled the roost. Look for Part 2 on Jan. 2.

Bruce Hoard is executive editor of SNW Online.

Copyright © 2005 IDG Communications, Inc.

7 inconvenient truths about the hybrid work trend
Shop Tech Products at Amazon