May 13, 2005
IBM has announced the DS4800, a storage product designed to compete in the mid-tier market with EMCís CX700 and HPís HP EVA 5000. The new storage product can operate in heterogeneous environments including AIX, HPUX, Solaris, Windows, and variations of Linux for both Intel and Power-based systems. The DS4800 also is supported by clustering software from a variety of vendors and will support IBMís new TotalStorage SAN Volume Controller and IBMís latest version of virtualization software. The new product will start shipping immediately and will be priced starting at $54,000.
With data creation doubling on a year-to-year basis, it is no wonder that more and more mid-tier enterprises (along with smaller companies) are seeking more storage capacity with increasing urgency. The pressure to do so comes not only from the demands of the raw amounts of data, but also from increasing requirements to store and manage that data. Such requirements are driven both legislatively and by market demands, as many mid-tier enterprises find themselves having data management decisions dictated by customer requirements. To not add storage capacity is to not do business, in many cases.
But as the demand for storage increases, the requirement for more complete offerings is growing as well. While IBM touted the speeds and feeds of the DS4800, we would argue the real value proposition is in offering not just hardware, but an actual solution package. The heterogeneous environment support is important, as is the variety of clustering options. The virtualization capability is one of the more powerful elements of this offering and one that truly differentiates this product from a mere box on which things can be stored. IBMís movement in this direction is one that we hope storage vendors en masse will continue to pursue. Storage hardware is nice; storage solutions meet existing market needs more effectively.
EMC has announced a four-node version of its Centera product which, when launched three years ago, was the first product specifically aimed at Content Addressed Storage (CAS) opportunity. CAS focuses on archiving of content that isnít going to change, such as medical records, finished documents, or cancelled checks. The new configuration scales to 2.2TB of storage, and has storage on access node capabilities whereas previous versions of Centera had one access node and the other nodes were for storage. New technology EMC has incorporated allows customers to put storage on both nodes. Incidentally, this technology is available for older systems with the newest version of the operating system. EMC has sold over 30PB of Centera storage now, with over 400 software partners. EMC estimates that at least 40% of customers are using Centera for compliance-related archiving.
While EMC was a market leader in the CAS space, they have been laggards in driving the product to the SMB space. The product has done well in its larger eight-node and sixteen-node versions, but those were too large and expensive for mid-market customers. Mid-market companies need to archive documents in the same ways that their larger counterparts do; they usually want the same kinds of technology, but in smaller or more manageable packages. Competitors such as HP and StorageTek have attempted to fill the gap with smaller, more affordable versions, and EMC should be able to grab more of the low-end market now with this new device. Server vendors, driven in part by the introduction of Linux, discovered a couple of years ago that the mid-market wanted products that were equally well-featured, and that it was degree rather than kind that mattered to them. The result has been server systems that are similarly featured from high-end to low-end, coming out fairly quickly after the initial launch of new technology. The storage industry has been slower to understand this. We can only hope that in the future vendors will be quicker to launch mid-market versions, as three years is a long time to wait for a product that is essentially the same, just smaller.
CAS as a market is a relatively new idea that is definitely here to stay. The need for archiving separate from backup is not yet well understood, but more companies are indicating the need to have both. We often believe that the storage industry continues to experience growth despite vendorsí collective marketing efforts rather than because of it, since the proper care and feeding of corporate data is still more art than science and storage vendors are still heavily oriented toward products rather than solutions. However, business changes are driving this need and to their credit, vendors such as EMC are seriously reworking their marketing to help customers understand how their products can be used to solve these business challenges through better messaging and packaging of solutions rather than merely offering lists of products of various vintage. Along these lines, the notion of a standalone archiving package is useful, particularly for smaller organizations or those with limited needs. However, as archiving becomes more popular, files earmarked for CAS may ultimately be only one descriptor of a greater document management strategy for corporations, and like many other IT advances, archives may become virtual over time. This means that the inherent value of Centera is not in the hardware but in the software, and that ultimately, customers will care less about what hardware is used and more about the management and security features the software provides. We hope that EMC will continue to drive CAS forward by learning to depend less on the underlying platform for Centera and instead focus more attention on driving the software to meet the greatest range of customerís architectural needs.
IBM announced this week that it has acquired Gluecode Software, a provider of software and support services for open source application infrastructure software for an undisclosed purchase price. Gluecode has used a download-for-free and pay-for-support subscription revenue model, one IBM will follow offering three different levels of support: Basic, Professional, and Premium for $500, $1,500, and $2,500 respectively. The Gluecode application server is based on the Apache Geronimo application server and† the J2EE development environment. IBM said the acquisition allows the company to reach further down into the market with Gluecode products as a development platform for SMBs and departmental projects.
IBM is positioning the Gluecode offering as a sort of Express version of its WebSphere Application Server Express, which costs more than all but the Premium version of Gluecodeís product. Company officials said they felt that offering a free initial copy of the product was likely to put it in the hands of more developers working on smaller scale projects and would allow them to begin the process of developing on the J2EE platform and in an open source environment more easily as well as developing applications that will run on the WebSphere architecture.
As we have noted repeatedly, IBMís commitment to the SMB market is company-wide and substantially deep. Adding Gluecode as a means to offer SMBs a less expensive and less complex means to begin building applications that can grow and mature along with an expanding IT ecosystem makes a great deal of sense. As developers build applications on Gluecode, they in essence are starting down the path that allows them to move up the food chain to build on and with WebSphere Express offerings and even up the enterprise calls offerings of the WebSphere line itself. While some may argue that such paths will be taken only rarely, we believe the annual doubling of data retention being experienced these days, along with ongoing complexity of regulations concerning that data, will force many companies to begin upgrading their infrastructure with more and more sophisticated products. IBM is providing developers and SMBs with an opportunity to get on that migratory path with a minimum of muss and fuss and in a way that gives those Gluecode users comfort that as they need to grow, they can do so with a relative lack of pain.
Sun Microsystems announced this week that it has entered into a definitive agreement to purchase all intellectual property rights of Procom Technology, Inc. relating to its NAS offerings in an all-cash transaction valued at approximately $50 million. Procom and Sun have had a software licensing agreement since April 2004 that resulted in Procom technology being embedded within the Sun StorEdge 5000 Family of NAS Appliances. With this purchase, Sun will own the intellectual property rights relating to this technology and gain additional engineering expertise. The transaction is subject to customary closing conditions, including regulatory approvals and the approval of Procom's shareholders, and is expected to close in June 2005.
It is becoming clear that the next move will be a flurry of storage-related announcements from the big boys. What is interesting about this announcement is not so much the actual company or technology being purchased but rather the strategy and market posturing revealed. A $50 million purchase is pocket change for Sun, even as the candlepower of the Copernican juggernaut is more subdued than in mightier times. However, the lessened stature of Sun is one reason this acquisition makes a great deal of sense. Sun in the last couple of years had learned that its SPARC+Solaris mantra was increasingly irrelevant in the lower echelon of the marketplace. The company has made notable strides in reestablishing its relevancy in this market segment through Opteron servers, creative licensing, and financing models, and acceptance that Linux on x86 servers is a valid and even respectable solution for many. While this change of heart is laudable, especially for the SMB marketplace, appropriately scaled storage solutions were few, which left the company at a competitive disadvantage in offering complete server and storage solutions to SMBs.
The solution-focused approach being demanded in the marketplace represents both a challenge and opportunity. For those with a comprehensive solution, it rings of low hanging fruit; for others, it demonstrates a hole in the product offering. The OEM deal Sun cut with Procom helped Sun offer more solutions to the lower end of the market, and thus access to this technology became in Sunís strategic interest. Given the rate at which larger players have been scooping up the OEM suppliers of competitors, it seems that Sun decided to act defensively and strategically in one fell swoop. Buying Procom keeps it out of the hands of competitors while also bolstering Sunís ability to generate additional NAS offerings, and possibly at lower prices. Given the incessant growth of data stored and burdening compliance and governance requirements, having a solid SMB NAS play could be a very helpful arrow in Sunís quiver that could help the company build future NAS and next-generation file-based storage systems more rapidly than it could before. Of course, having a product to sell and getting people to buy are two different things. Perhaps Sun will apply some of its creative thinking to packaging and pricing NAS as well to helping drive its acceptance in the marketplace. While this purchase in and of itself will not single handedly return Sun to its glory days, it does once again demonstrate the tenacity of the Santa Clara player, and its desire to play in all aspects of the market, not just the lofty high end.
Novell announced this week its acquisition of Immunix, provider of host-based Linux application security solutions. Immunixís AppArmor product, protecting both the operating system and applications from external or internal attacks, viruses, and other malicious exploits, will be branded Novell AppArmor Powered by Immunix. AppArmor seeks to reduce the need for urgent, emergency-reactive patching and instead offers patching as a scheduled maintenance activity. AppArmor offers predefined application security policies for common applications such as Web, email, and remote login, as well as wizards to facilitate the deployment of custom security policies. AppArmorís built-in reporting and alerting allows IT to ensure security policy compliance, with realtime notification of any attempted breeches. The financial terms of the deal were not disclosed at the announcement.
For Novell the acquisition of Immunix is complementary to its Linux portfolio, and is synergistic with the push of SuSE Linux Enterprise Server into the datacenter. As Linux gains momentum and popularity, its exploit exposure increases and datacenter success depends on effective neutralization of these threats. By joining forces with Novell, Immunix can quickly and cost-effectively extend its global market reach and gain access to a robust pool of expansion capital. Unlike signature- or anomaly-based detection/protection mechanisms, host-based IPS is effective at stopping zero-day attacks, requires less maintenance, and enables a more stable patch management environment. This functionality should prove attractive enough to induce enterprise IT to incorporate host bases IPS into their security umbrella.
The Immunix technology faces a range of network and host-based IPS competitors such as Bodacion, Determina, Captus Networks, and the Barrier Group ó mostly smaller, early-stage companies ó as well as the more traditional intrusion detection/prevention vendors such as ISS, Symantec, and McAfee. For the smaller IPS vendors unique technology will probably be insufficient to ensure survival. As in the case of Immunix, it chose to seek partnerships, alliances, or acquisition with or by larger players with a global presence. Also enterprises increasingly are seeking integrated umbrella security solutions provided by a single source, rather than separate stand-alone point solutions. And, security is becoming part of the network fabric, spanning all network connection, devices, and servers. Larger multi-solution providers like Symantec, Cisco, and IBM are in the best position to capitalize on the above trends. Novell, with its anchor position in the Linux market and with a global presence, is in a good position to win share in the IPS market. The characteristics of the Linux market provide vendors with multiple IPS strategy alternatives. At a product level, for instance, host IPS could be tightly integrated with the operating system, or offered as an appliance on the network. In our opinion a network-based, component, or appliance IPS approach will be favored by enterprise IT, since it has the potential to offer lower-cost solutions with greater flexibility. On the vendor side, the Novell-Immunix strategy appears to favor a loosely integrated approach to the IPS, since this potentially offers the greatest revenue opportunity for the parties involved.