November 30, 2007
HP has recently announced a collection of virtualization and power management technologies that are designed to help customers streamline IT operations, realize cost savings, increase flexibility, and improve energy efficiency. The HP Virtual Connect Enterprise Manager extends the capability of HP BladeSystem Virtual Connect technology to all blade enclosures in a data center. The new Enterprise Manager enables IT administrators to manage and control these connections across 100 c-Class enclosures, or up to 1,600 blade servers, from a single console. The new HP Server Migration Pack Universal Edition combines virtual and physical migrations into a single tool to speed migration time of HP ProLiant and BladeSystem servers. The Management Pack provides central management with expected support of Citrix XenServer, Microsoft virtual machines, Oracle VM, and VMware. HP PolyServe Software for Microsoft SQL Server consolidates large SQL environments onto a single cluster so customers can manage all instances at once, freely add and recover multiple instances, and roll out business applications more quickly while improving reliability. HP stated that it plans to integrate the Opsware Automation Platform with its existing management solutions for future automation of data center capabilities across the full software stack.
Also included in the announcement is the new HP Power Distribution Rack, which controls three-phase power distribution across a row of server racks. The offering allows IT managers to connect to power once across a row of server racks and adapt power distribution as needed; prevent overloads and resolve problems quickly with HP Thermal Logic technology; and reduce cabling complexity with one set of input cables to the end of a row and short power drops to each rack. According to the company, the new HP Rackmountable Parallel 3 Phase UPS provides the highest level of power protection from HP and dissipates less than half as much heat as competitive offerings. It enables attached servers to save all work in progress and initiate a shutdown in the event of power loss, and restores it with Thermal Logic power policies.
The new technologies in this announcement are actually rather useful for the intended audience. The Virtual Connect Enterprise Manager offers a considerable range of control over many hundreds of servers from a single console, which is the kind of control that IT managers tend to favor. Likewise, the Server Migration Pack Universal Edition offers another desirable capability, namely the ability to handle virtual and physical servers in the same fashion with the same tool. Further bolstering its power-aware position, HP Power Distribution Rack brings enhanced flexibility to IT personnel who are seeking to place power where it is needed, but without over provisioning or otherwise wasting the precious resource. According to the company, the new 3 Phase UPS can save $1,000+ annually in power and cooling costs in the 12-kilowatt rack-mount model and $6,000+ for the 60-kilowatt row-level configuration. These numbers are impressive and are worthy of PR attention Overall, we see these new offerings as valuable assets for data center managers. So what’s the rub?
What’s troubling about this announcement is not the technology, nor the vision of the company, but rather the seeming need by the company to promote a larger-than-necessary aura around everything HP. These announcements were part of a long press release from HP that covered a litany of existing and new product offerings. At first glance, one might be tempted to view this as merely undifferentiated verbiage about blades and energy efficiency in the data center. However, there were new product details buried in the announcement, and it is unfortunate that clarity may have been sacrificed by the company’s perceived need to be seen as larger and omnipotent than the competition.
HP clearly remains an engineering-focused firm, and arguably this is one of its strengths. So too is its ability to create well rounded solutions that transcend a mere collection of point products. Yet at times the company’s messaging can become so focused on bolstering its position (declaring itself King of the Hill) that it loses the ability to let its products speak for themselves. In an era where the totality of the product, solution, and services is paramount, at first glance this messaging approach would seem sound; however, it can also come across as obsessive, if not paranoid. To our way of thinking, removing the aura of grandeur often results in a grander and more honest position than the puffed up alternative. Further, the clarity of what is being offered, and its value, is enhanced. HP has extensive, credible credentials in the blade and energy efficiency arena. We only wish that the firm would at times realize it is often best to state simply and plainly what’s new and exciting and leave the extracurricular PR fluff behind. This would make it easier for their customers and business partners to immediately glean the considerable value the company can impart, while also delivering a more confident and down-to-home feeling to an important player in the IT marketplace.
IBM has recently announced BladeCenter Open Fabric Manager (OFM), which provides I/O virtualization through an open architecture that supports a range of Ethernet and Fibre Channel technologies from vendors including Blade Network Technologies, Brocade, Cisco, Emulex, NetXen, and QLogic. IBM BladeCenter Open Fabric's use of open technologies and support for third-party management solutions allows clients to leverage existing switch technology. The core I/O virtualization technology is also unique in that it manages up to 100 chassis from a single management console. IBM also announced a statement of direction that POWER6 processor-based blades will support i5/OS applications, which have traditionally been supported on the System i server platform. IBM BladeCenter Open Fabric Manager is priced at $1,499 and IBM BladeCenter Open Fabric Manager–Advanced Upgrade is priced at $1,999 with availability slated for both products in December. Until the end of the year, IBM is running a promotion in the U.S. that offers the product at a discount of $500 to $1000 per chassis off the list price.
This announcement is interesting in part because is yet another piece in the virtualization puzzle, but also in the way it differs from competitive approaches such as HP’s Virtual Connect with respect to the integration of blade chassis into the larger datacenter context. Through its support for most of the standard and common interconnect technologies in the datacenter, OFM offers organizations a straightforward method by which to interconnect their new investments in blade chassis into the existing datacenter architecture. Although many, including IBM, HP, and others, have stated that the future of the datacenter is blades (and we would be hard-pressed to disagree), the reality of the present is that there are huge, functioning, IT investments not bladed by nature. As part of the eventual transition to blade and chassis future, it is imperative that organizations not inadvertently create new silos of IT with blades that are unable to interconnect with existing server, storage, and other network resources. In addition, organizations could create blade failover pools and hence support failover scenarios whereby blade and virtual LANs are reconfigured on the fly to meet the HA requirements for mission critical applications. Addressing this need and other IT needs is part of market drivers that has caused vendors to create solutions such as OFM and Virtual Connect in the first place.
With this in mind, part of what we find compelling about OFM over other solutions is that organizations can continue to use the switching fabrics that they have already deployed without necessitating the purchase of new, redundant, and proprietary interconnects to order to integrate fully their blade chassis with the existing infrastructure. This is one way in which IBM’s solution differentiates itself from Virtual Connect and other interconnect strategies. While one can logically argue that in a greenfield deployment the investment in a specific switch is a moot point, the reality is that there are few, if any, greenfield IT deployments these days. The focus on blades and consolidation is about virtualization of as many IT components as possible, at the chassis, blade, slot, and I/O interface level. From these pools of resources, one can orchestrate the virtual infrastructure necessary to support a workload portfolio and with appropriate management software dynamically alter the physical mapping as needed.
Overall, we are pleased with the decoupled nature of OFM, whereby the software and management functions are discrete from any support switching mechanism. Not only does this free the customer from specific product lock-in, it provides a greater leverage of existing switching investments in the organization. As such, this delivers additional value to the organization on top of its blade and legacy infrastructures in a truly incremental and non-redundant fashion.
CA Inc. and HCL Technologies have announced an agreement to establish a strategic partnership in which HCL will assume all research and product development connected with CA's threat management security business, while CA will retain all sales and marketing functions. The goal of the strategic partnership is to grow CA's threat management business by combining the strengths of both organizations. HCL and CA will achieve goal alignment and financial targets through revenue sharing. The annual revenue of CA’s threat management security business is in excess of U.S. $100 million. The partnership covers all threat management products, which include antivirus, anti-spyware, integrated threat manager, host-based intrusion prevention system, secure content manager, Internet security suite, anti-spam, and firewall. HCL will be responsible for research, engineering, architecture, technical support, technical writing, and quality assurance. CA’s threat management products will continue to be sold exclusively under the CA brand and through several routes to market with a growing emphasis on channel partners. The partnership is expected to become operational by year-end, following the signing of a definitive agreement.
While there is a global economy, there is still a parochial approach to security research and product development. Notwithstanding the likelihood of costs being lower in Noida, India than they are in Islandia, Long Island, New York, national governments have always viewed offshore development, influence, ownership, or control as a red flag. This jaundiced view is not necessarily due to the quality of the product development, nor to the ultimate features and benefits; it relates more to the perception that adversaries will be able to exploit the offshore nature of the development to their advantage. This action is very likely to torpedo much of CA’s efforts in defense departments outside of India. If, however, the company has decided to expand its defense marketing into Asia, then the move might have less of an overall negative impact.
To some people $100 million would seem like quite a significant sum; however, in the context of an organization with revenue north of $1 billion per quarter, the amount is rather infinitesimal. Overall, we view this as an early intelligence indicator that CA is not serious about security and is likely to divert its efforts elsewhere. It would appear that CA has made the internal decision to take some of its eggs out of the security basket and concentrate in other market segments where its internal development efforts are likely to bear more fruit. Given Symantec’s recent lackluster financial performance and the apparent ascendancy of EMC in the security market, CA likely hopes that its sales force and key SI partners will attack segments where CA brings value and faces less competition than it does in security.