Storage Pipeline: Review: Out of the Abyss

Enter storage-resource management, an approach Legato (now a division of EMC), Tek-Tools, Veritas and several other vendors tout with their various suites and frameworks. SRM comprises a collection of software tools for discovering where storage is located and how its capacity is being used. The more sophisticated products monitor storage hardware capacity utilization and other storage-related software processes, such as data replication, volume management and backup, to provide alerts about potential problems in near real time.

SRM Timeline

At the turn of the century, SRM was touted as a silver bullet. The software was to enable fewer managers to oversee more storage and to provide ample warning so that steps could be taken to solve problems before they created downtime. By 2001, however, SRM came under attack for being too platform-focused to address the real goals of storage management. Critics said you couldn't manage storage by taking the temperature of various infrastructure components; you needed to look at what effect storage was having on application performance. BMC Software spearheaded efforts to deliver an application-centric, policy-based approach to storage management that could drive labor costs out of storage investments.

Many SRM vendors seized upon this idea to create complex products typified by an automated, application-aware storage-management software stack. These stacks comprised a policy engine riding atop a storage-virtualization layer (virtual volume managers for block storage and global namespace products for file systems) that automated certain tasks, such as growing virtual volumes to meet burgeoning application data storage requirements, and improved single manager productivity. Work continues in this area as many brand-name management-software vendors begin porting their technology into storage-switching platforms or multifunction storage-management appliances.

But the story doesn't end there. Over the past year, vendors have started chanting yet another mantra: data life-cycle management. In this approach, information about applications and storage platform capabilities and costs is used to create a knowledge base that lets policy engines create rules for automating the smooth migration of data across different storage devices. Instead of focusing on the application or storage infrastructure solely, this model makes data its centerpiece.

According to EMC's Legato and others embracing this vision, the data life-cycle management revolution will be more effective at automating management and reducing costs than any hardware or software innovation to date. It will create the optimal storage-utility infrastructure, in which capacity is allocated dynamically and automatically to applications that need it, and data will move from platform to platform automatically, based on its usage characteristics, retention requirements, platform costs and other factors.

But there are limitations in the life-cycle management products now emerging. One of the most nagging is the absence of an open standard for equipping data with a self-descriptive header that would identify its requirements or originating application. This header would help automated management tools move the data around. Currently, all the approaches for data self-description are proprietary and limited. However, work is under way at both NASA's Jet Propulsion Laboratory and the International Standards Organization to create a standardized naming convention (see www.iso.org/iso/en/commcentre/pdf/Data0001.pdf).

The difficulties don't end with the header issue, though. Life-cycle management confronts much the same set of problems that every other management scheme faces, except point management. Specifically, hardware products have not been designed for common management within a heterogeneous environment. Management-software vendors are at the mercy of hardware vendors for access to their APIs and other configuration and monitoring controls, and many hardware vendors simply are not interested in supporting anything that would enable competitive products to share space within the customer's shop.

The storage utility remains an airy vision without much substance--more "marketecture" than architecture--and storage management in a heterogeneous environment continues to be a work in progress. To gauge the progress that has been made in delivering the various functions that make up the utility storage-management stack (see "SRM Stack," at right), Storage Pipeline created an RFI featuring a hypothetical company with some real-world storage configurations. We sent the RFI in August to about 20 of the industry's leading storage-management software vendors. Fictitious Minuteman Mortgage presented its storage-management conundrum, and three vendors, Computer Associates International, Fujitsu Software Technology Corp. (Softek) and Storability, responded with in-depth solutions.

In the end, it was a tough call. In fact, all three solutions earned a B+ (see our Report Card). The solution from Storability scored solidly for well-designed business views, intuitive operability, broad standards support and well-integrated core components. But its solution overall wasn't the best fit for Minuteman Mortgage.

Fujitsu Softek's solution also had some impressive strengths. It showed strong policy support, excellent environment support and strong topology discovery and visualization.

All three were viable, and all three vendors are qualified to do the job. But Computer Associates' BrightStor proposal--with its emphasis on policy-driven data movement, without a requirement, at least initially, to add unfamiliar technology layers like virtualization--stood out. We agreed with Computer Associates that improvements in capacity allocation and utilization efficiency could be made quickly and permanently if Minuteman simply implemented the disciplined management of its storage.

Our complete RFI and the respondents' proposals can be found online here.

To summarize, Minuteman Mortgage sought comprehensive management for its storage infrastructure, which contained a broad mix of platforms and topologies. With 30 TB of heterogeneous storage, primarily in direct-attached configurations, and a Fibre Channel SAN fabric, obtained through an acquisition of another company and which had reached its maximum node capacity, Minuteman's management problems were multiplying.To make matters worse, it had maxed out the node capacity on the SAN.

Minuteman's problems also featured a few extra twists that we hoped would help differentiate the responses we received. For example, some of Minuteman's storage was shared within a server cluster, and as some companies have discovered, many storage-management products do not handle this configuration well: They treat as unique the storage that each server sees. This, in turn, introduces a flaw in capacity and usage calculations, because each server in the cluster sees the same storage.

Minuteman also used many storage platforms capable of native management solely via a self-articulated Web page. We wanted to see how vendors would consolidate this data so that managing storage wouldn't feel like surfing the Web.

Our savvy IT manager at Minuteman also sought a solution that wouldn't charge him for functionality he didn't need or use. This is a common complaint; with some SRM suites, managers are forced to license many modules they don't use to obtain the few that they do.

Finally, though our fictional IT manager was open to suggestions about infrastructure revisions that might improve his environment, he preferred a storage-management recommendation that wouldn't require a forklift upgrade of his infrastructure. With many storage-management products, limitations in device support (array, switch or server platforms manageable via the product) or topology support (for direct-attached, network-attached or SAN-attached) may be unclear from vendor-supplied literature, forcing the customer to shelve the software or make costly infrastructure modifications.

A perusal of the responses from the three vendors revealed common themes and messages regarding some principles of storage management. For example, all three vendors stressed support for heterogeneous storage platforms, server operating environments and topologies. In CA's case, we had to visit a Web list to track the constantly changing set of target devices, firmware revisions and host bus adapters, for instance, supported in its BrightStor suite. Fujitsu Softek promised to support whatever servers and storage platforms users chose to deploy. Storability's bid listed the platforms it supports--a brave move but perhaps foolhardy because it revealed its lack of support for some of the devices deployed by Minuteman. Storability provided no explanation for how it would support nonlisted devices. Only Computer Associates showed clear support for all the platforms listed in our RFI.

The three vendors also rallied around the battle cry of automated provisioning. There were subtle differences, however. For Fujitsu Softek, whose solution set includes an in-band virtualization engine (a customized version of DataCore Software's SANsymphony product), provisioning seemed to mean creating a pool of unallocated physical storage capacity from all managed arrays, whether connected in direct-attached storage, network-attached storage or SAN topologies, and then parsing it out as virtual volumes that could be scaled on command behind applications.

Storability also suggested this approach. Its proposal referenced "an optional Catalyst module" that might be used in connection with its Global Storage Manager SRM product to create scalable virtual volumes via array "LUN masking." Storability would also support certain third-party LUN (logical unit number) masking products or virtual volume managers, like Veritas Volume Manager, if they were already in use at Minuteman.

Declarations about the necessity of storage virtualization for effective provisioning were notably absent from the Computer Associates proposal, however. In response to our question about how provisioning would be automated, CA instead discussed data migration. The vendor seemed to be treading a thin line between conventional wisdom about the need for virtualization and what its own experience had taught it about the amount of storage lost to poor management in general. Although CA refrained from making explicit statements, its response implied that provisioning in most shops wouldn't be nearly the pain that it has become if a system of policy-based data migration were applied rigorously and capacity managed properly in the first place.

One might interpret CA's response as an effort to deflect attention from the fact that the vendor has no virtualization solution of its own. However, it's reasonable to assume that CA's BrightStor will support most any third-party virtualization approach, so the decision not to mention virtualization in the proposal seems deliberate. It is more likely a reflection of the vendor's belief that disciplined and informed data management--including policies and procedures for culling junk and stale data from primary storage--is a more effective route to capacity allocation efficiency than is the addition of complex virtualization technology.

Although Minuteman's problems might be addressed in the short term by a dollop of virtualization technology, such a strategy might not solve its capacity conundrum in the long run. In the final analysis, we preferred a strategy focused on policy-based provisioning without virtualization.

If virtualization were determined to be required at a later date, a best-of-breed product could be selected and deployed with full support from the CA product.

Computer Associates also parted company from the other vendors by making it clear that Minuteman needed to begin evolving its infrastructure toward a SAN. Given Minuteman's heavy emphasis on database-centric applications and its experience with the vicissitudes of Fibre Channel SAN technology, the company seemed to possess the hard-bitten realism that might make it a good candidate for expanded use of the topology. This would also facilitate CA's preferred enterprise-focused data protection and data life-cycle management solution by providing a multipathing off-LAN interconnection between storage platforms.

There were weaknesses in CA's proposal. For one, the company was fuzzy on the details of its support for capacity reporting on server-clustered storage. Both Storability and Fujitsu Softek gave explicit answers that demonstrated they had encountered and understood the problem. Computer Associates also demonstrated a lack of willingness to commit to a final price for the solution--or even a ballpark estimate--without a fact-finding mission. By our calculations, CA's solution would cost in the neighborhood of $52,000 for software, plus $5,000 per year for maintenance. The price for implementation services, which could require several weeks of on-site consulting and training, was undisclosed. Its bid also omitted pricing for server hardware that Minuteman would have to buy to host BrightStor Portal, SAN Designer, SAN Manager and Enterprise Backup software.

Taking these costs into consideration, the price would be near or greater than the price of the Storability solution, which totaled $249,120, with maintenance, installation and training. Of course, Storability's pricing does not include some prerequisites, like deployment of Microsoft's Internet Information Server and a SQL database. It was also unclear how much additional cost would accrue to Minuteman's election to use the Catalyst Module with Storability's Global Storage Manager solution.

In the final analysis, Computer Associates edged out Storability because of the latter vendor's approach to the presentation of the storage infrastructure on its console. We wanted to see storage presented from the application or business process perspective to simplify troubleshooting and error resolution. Storability instead presented storage according to the functions performed by an administrator. Although useful when providing remote storage services to a client with many data centers, this function-oriented approach seemed less applicable to Minuteman's situation.

Compared to those solutions and prices, Fujitsu Softek looked like the "little engine that could." Its solution was the fastest to deploy (on paper at least), at two to five days--priced at $2,500 per day with implementation support services. And it was the lowest priced from a licensing perspective, at $104,900 plus annual maintenance of 18 percent. We believe Fujitsu Softek's claim that the customer would start to realize value as soon as the solution was up and running. Computer Associates narrowly edged out Fujitsu Softek's solution for several reasons, including Softek's lack of quantification of the cost or time required to build "action sets," or policies, for managing the Minuteman infrastructure, which we viewed as the heart of any real solution.

Fujitsu Softek offered a technological quick fix to Minuteman's management boggle and at a comparatively low cost. Although it's still a young company with a young product that lacks seamless component integration, we believe that CTO Nick Tabellion can steer the products to fruition over time, as he did SMS (Systems Managed Storage) at IBM. SMS is the granddaddy of open-systems SRM.

What nagged at us, however, was the lack of pedigree in the Fujitsu Softek product. Though the proposal suggested that action sets could be developed to automate provisioning functions in the solution, the document simply did not convey the level of experience offered in the much lengthier and more heavily diagrammed Computer Associates proposal. We were also concerned that Fujitsu Softek's proposal emphasized the full faith and credit of Fujitsu Ltd. as a guarantee of the company's future and ability to execute on its product road map. Yet Softek is working to separate itself from its parent company. Softek would have been better advised to stick with its strengths--Tabellion and his team--to make the case for its technical solvency than to lean on the Fujitsu connection.

So we concluded that Computer Associates BrightStor was the best fit. Our decision came down to three factors:

1. Computer Associates offered a conservative and carefully reasoned solution to the needs of Minuteman, a financial firm subject to a broad range of explicit and demanding government regulations with respect to consumer data. The proposal suggested that disciplined policies were more important than gee-whiz technology, and we agree.

2. The BrightStor proposal emphasized the need to understand business processes and to map data access, protection and retention characteristics to platforms via policies before making decisions about new hardware, deployment topologies or appropriate application of technologies like virtualization. So much emphasis was placed on situation-assessment tasks that must be undertaken before deployment of the BrightStor software components that the proposal appeared more like a consulting engagement than a product sales pitch. Usually, this would be a negative, but in the case of storage management, a heavy emphasis on situation assessment, gap analysis and pre-implementation design is required if the solution is to deliver value.

In this respect, CA's proposal reflected hard lessons learned in the company's past efforts to sell its UniCenter enterprise-management solutions and a real understanding of business-technology nexus.

Bottom line: Computer Associates didn't view storage management technology deficits as the root cause of Minuteman's problem, but rather the lack of a management strategy. Although the customer might opt for a quick fix or a technology silver bullet, Computer Associates was willing to risk losing the bid with the proposition that the problem would not be truly resolved until IT was harnessed to the business process via policies based on business reality. We could find no flaw in that assertion. Storage management software is at best an application tool kit that must be custom-built for each implementation.

3. The final vote for Computer Associates came down to integration. Components of the BrightStor suite are more tightly integrated than are the components of the Fujitsu Softek offering. CA has a knowledge base of best practices that reflects its long tenure in enterprise management and a common services layer that provides a kind of glue between its BrightStor components and their management targets. Moreover, storage management via BrightStor could well become a stepping stone for Minuteman into the bigger world of enterprise management--evolving from a utility storage solution to a utility computing solution.

In our view, Computer Associates wins this round, with both Storability and Fujitsu Softek close. In the real world, the decision might come out another way. Corporate culture might cast the deciding vote. Fujitsu Softek offers a tactical and effective fix to an immediate problem with minimal disruption to normal operating procedures. Some might see it as the path of least resistance. However, in organizations willing to invest in a long-term solution, Computer Associates would probably have the greater appeal.

Jon William Toigo is CEO of storage consultancy Toigo Partners International, founder and chairman of the Data Management Institute, and author of 13 books, including Disaster Recovery Planning: Preparing for the Unthinkable (Pearson Education, 2002) and The Holy Grail of Network Storage Management (Prentice Hall PTR, 2003). Write to him at [email protected].• BrightStor Storage Management Solution, as configured for this test: BrightStor SAN Manager, BrightStor Storage Resource Manager, BrightStor SAN Designer, BrightStor Enterprise Backup and BrightStor. Computer Associates International Inc., (631) 342-6000, (800) 225-5224. www.ca.com

• Softek Storage Manager, as configured for this test: Softek Storage Manager (with SANView), Softek Storage Manager Space Optimizer Agent, Softek Storage Provisioner. Fujitsu Softek, (877) 887-4562. www.softek.com

• Global Storage Manager. Storability Inc., (508) 229-1700, (866) STRGMGT. www.storability.com

• "Building a Storage Management Strategy in 2003"

• "The Myths of SRM"

• "Network Storage Resource Management Secrets From the IT Departments of Major Corporations"

• "Calculating the Business Value of Storage Resource Management Software"

• "Storage Resource Management: A Practitioner's Approach"

R E V I E W

Storage Resource Management

Sorry,
your browser
is not Java
enabled

Welcome to

NETWORK COMPUTING's Interactive Report Card, v2. To launch it, click on the Interactive Report Card ® icon

above. The program components take a few moments to load.

Once launched, enter your own product feature weights and click the Recalc button. The Interactive Report Card ® will re-sort (and re-grade!) the products based on the new category weights you entered.

Click here for more information about our Interactive Report Card ®.