The Enterprise Server Impact Of Intel/AMD Settlement Remains To Be Seen

As an infrastructure and operations professional, it’s hard not to like the terms of the settlement between Intel and AMD but the implications for your buying strategy and roadmap aren’t that clear yet. The rules of engagement spelled out in the Intel 8K make it clear that server vendors are free to select the CPUs that best fit their design goals and to carry as wide a portfolio of AMD-based servers as they feel the market demands. Server vendors can also market AMD to their heart’s content without fear of retribution from Intel (if the accusations from the past are true — something Intel isn’t admitting through this settlement). It also means that the two companies can stop aiming so many legal weapons at each other and redirect those funds and management brain cells to accelerating innovation. The $1.25B influx to AMD will be a very welcome shot in the R&D arm, too.
That’s all goodness that may lead to greater adoption of AMD-based servers in the future by enterprises. It doesn’t mean we can expect equivalency in performance or purpose between them, though. All it does is spell out the rules of engagement for fair play.

It also doesn’t ensure full compatibility between AMD and Intel processors and this is where one question in particular lies unanswered.

According to the 8K, Section 2.3 on Technical impairments:
“Intel shall not include any Artificial Performance Impairment in any Intel product…mean[ing] an affirmative engineering or design action by Intel (but not a failure to act) that (i) degrades the performance or operation of a Specified AMD product, (ii) is not a consequence of an Intel Product Benefit and (iii) is made intentionally to degrade the performance or operation of a Specified AMD Product. For purposes of this Section 2.3, “Product Benefit” shall mean any benefit, advantage, or improvement in terms of performance, operation, price, cost, manufacturability, reliability, compatibility, or ability to operate or enhance the operation of another product.”

The key word here is “compatibility.” Case in point: virtualization live migration. Intel VT, the set of instruction sets in the Xeon CPU family that helps with the performance of virtual infrastructures and enables live migrations from server to server are not compatible with AMD-V technologies that do the same thing (roughly). What this means today is that with Intel’s implementation you can’t live migrate from an Intel-based server to an AMD-based server. AMD-V actually lets you migrate from AMD to Intel, but you can’t move back. As a result of this you are probably building out your server virtualization pool today purely with Intel-based servers.

So here’s the question: “Is the incompatibility of virtual infrastructure implementations an Artificial Performance Impairment or simply different approaches to delivering Product Benefit?”

The Intel 8K then states:
“In no circumstances shall this Section 2.3 impose or be construed to impose any obligation on
Intel to (i) take any act that would provide a Product Benefit to any AMD or other non-Intel product,
either when such AMD or non-Intel product is used alone or in combination with any other product, (ii)optimize any products for Specified AMD Products, or (iii) provide any technical information, documents, or know how to AMD.”

Hmm. This seems to state that Intel is under no obligation to address this particular example if the resolution provides a Product Benefit to AMD. And it further states in that same “(i)” that such Product Benefit includes cases of the products being used in combination – which describes the case of Intel and AMD servers in a shared virtual pool. It further states that Intel doesn’t have to provide any technical information, documents or know how to AMD through which they could take the initiative themselves to solve this problem.

If AMD can’t penetrate any enterprise’s virtual server pool because the feature of live migration doesn’t work across server families, that would seem to me to be a serious Artificial Market Impairment. But does that fit the terms of the settlement that state Artificial Performance Impairment must be proven? Hmm.

This is just one example of the types of questions and scenarios Intel and AMD’s lawyers and technical experts will have to apply to this agreement. And answers to these types of questions will be imperative for enterprises to understand as they can have significant impacts on buying behavior. 

Until these types of scenarios are played out the impact of this settlement on enterprise buyers remains unsettled.

By James Staten

Check out James' research

You can follow James on  here

CA Buys NetQoS: A New Network Management Juggernaut?

CA is a vendor that already enjoys a leading position in overall network management. Its 2005 acquisition of Concord, which brought along the assets of the previously acquired Aprisma, instantly moved CA from an also-ran to one of the clear leaders. Concord was good, and CA has an impressive track record of growing that business since the acquisition. Still, there were some weaknesses with regard to more advanced performance analysis.

On September 14, 2009, CA finally addressed these performance gaps by announcing its intent to acquire NetQoS for $200 million. Based in Austin, TX, NetQoS is one of those exciting small companies that proved there is a better approach to many of the challenges we face. It is one of the true innovators in performance management of both infrastructure and applications.

The combination of the two is a compelling proposition for network management. CA’s Spectrum is recognized as a leader in fault management, both in sheer installed base as well as technical merit. The eHealth side of the business (the original core Concord business) has been popular for performance monitoring and reporting, but the analysis from NetQoS is far superior. This is where the action now shifts in network management as well as other domains such as virtual servers, storage, and applications. The merged product line positions CA in a formidable position to capitalize on the growth areas of network management.

Network Management is a corner of the IT Management Software market that has been written off by many as boring and obsolete, yet we project that 2009 sales of network management products will be slightly less than $3.5B, or 17% of the overall IT Management Software market. This represents a 4% increase in raw spending over 2008. Granted, Forrester’s growth projection for 2009 in the broader IT management software market is 9.5%, so network management is lagging behind the overall market. Still, almost no IT spending is on the increase this year, so this is notable. Clearly, there is still demand. Network Management is hardly a dead market!

Certainly, many vendors are strong point players, but only a few can provide broad-based network monitoring and analysis for both fault and performance management, as well as configuration and change management. These powerhouses include:

• HP, which holds a longstanding leadership position centered around Network Node Manager and bolstered by the Network Automation product that came with Opsware (which it acquired with Rendition Networks).


• IBM, through the 2005 acquisition of Micromuse, gave it superb network discovery and fault management, along with a good set of performance monitoring capabilities. Prior to Micromuse, IBM’s Tivoli division was very weak in the network domain. Now it is clearly in the lead pack.


• EMC, the storage king that made its foray into the network with its 2005 Smarts acquisition. Smarts was, and still is, a phenomenal product for network fault management. In 2007, EMC added to its arsenal with the Voyence acquisition for NCCM. It still lacks a compelling performance-oriented story, but it is also among the leaders in network management.

We believe CA will augment much of its product line with NetQoS rather than replace it. There are obvious overlaps, but more opportunities to strengthen what already exists.

Is the new CA a potentially undisputed leader in network management? Possibly. The fault and performance combination is tough to beat. If a stronger network configuration and change management component existed in the CA portfolio, linked by its Unified Service Model and orchestrated by components of its Spectrum Automation Manager, the answer to the question would be a firm, “Yes.”

More importantly, the NetQoS technology promises a stronger position beyond network management. The need to gain better insight to performance, and especially application performance, is one of the most overwhelming issues still facing IT. If CA properly leverages the NetQoS portfolio, it can significantly bolster its server and application performance story. Many of the analytical algorithms extend beyond the network, so other management domains can benefit. NetQoS brings some attractive application performance management technology that should mesh with CA’s Wily product line. Here again there is some overlap that must be reconciled, although the two work in different ways and a hybrid that takes the best of both is very appealing.

NetQoS has been successful because it innovated in an area that most other vendors ignored. It is a relatively small company, but we expect it to have a big positive impact on CA and its customers.

By Glenn O'Donnell

Check out Glenn's research

You should follow me on  here

Will Advances In The Mac OS Tip The Scales For BYOPC?

(Reblogged from Natalie Lambert's personal blog I Brought My Own PC)

CNET recently reported that Apple’s upcoming Snow Leopard OS will support Microsoft Exchange Server 2007. Why is this important for BYOPC? Because until this release, BYOPC presumed some amount of virtualization (in most cases desktop virtualization) to support the killer app — email. According to the CNET article, “Improved Exchange support will be integrated into Mail, iCal and Address Book in Snow Leopard, which means email, calendar appointments, to-do lists and contacts from Outlook will be viewable on your personal calendar, mail and address books. It also allows things like dragging and dropping contacts into iCal to schedule meetings, and your Mac will be able to discover time conflicts between personal and work calendars and change the meeting time and location.”

So, with virtualization no longer needed to support the killer app on a Mac, organizations will need to rethink how they will support users that want to use Mac’s in their native state for corporate activities. However, this is a huge step forward for employees wanting to use Mac’s in the corporate environment…IT can no longer hang their hat on email not being supported. This opens a few questions for me:

• will IT shops have the proficiency to support the Mac?
• will email work well on a Mac and live up to expectations?
• will this follow a similar trajectory as the iPhone with execs just demanding it?
• what is the fate of VMware Fusion and Parallels?

Regardless, my advice is this: Walk down to your local Apple store and offer the Genius behind the counter a job!

By Natalie Lambert

Check out Natalie's research

You should follow me on  here

Cloud DR Services Are Real

There is a lot of hype surrounding cloud and I'm usually not one to join the hype but in the case of cloud-based backup and disaster recovery services (I'm trying to use the IT service continuity but it hasn't caught on yet), these service are available today and they address major pain points in IT operations and organizations of all sizes can leverage these services, not just small and medium businesses.

Storage-as-a-Service is relatively new. Today the main value proposition is as a cloud target for on-premise deployments of backup and archiving software. If you have a need to retain data for extended periods of time (1 year plus in most cases) tape is still the more cost effective option given it's low capital acquisition cost and removability. If you have long term data retention needs and you want to eliminate tape, that's where a cloud storage target comes in. Electronically vault that data to a storage-as-service provider who can store that data at cents per GB. You just can't beat the economies of scale these providers are able to achieve.

If you're a small business and you don't have the staff to implement and manage a backup solution or if you're an enterprise and you're looking for a PC backup or a remote office backup solution, I think it's worthwhile to compare the three year total cost of ownership of an on-premise solution versus backup-as-a-service.

If you lack a recovery data center or if you're looking for more a cost-effective solution for your Windows and Linux x86 server farms, I recommend that you evaluate new virtual recovery services. And for anyone who is about to renew their DR services contract, you should evaluate these new services before you resign for another three years of tape-based recovery to oversubscribed, shared IT hardware. These services fill the gap between DR services that rely on replication to dedicated IT hardware which are far too expensive for most organization and tape-based recovery to shared IT hardware, which while affordable, do not meet the recovery time and recovery point requirements of organizations from small to large and acoross industries.

And of course, if you're focused on the availability and recovery of email in particular, well there is a laundry list of service providers that focus cloud-based services for email from including continuity of service, recovery, archiving and if you wanted to, just move your email to the cloud period.

These services already have a toehold in the US but in order to expand into Canada and to Europe, these service providers are busily opening up data centers in these regions. In Canada and the EU, data privacy laws restrict the transfer of data containing personally identifiable information out of region or to a country that doesn't meet the standard of its own privacy laws.

Perhaps I'm too bullish on these services, perhaps organizations still have security concerns with cloud-based services but I think evey IT operations professional must now seriously consider these services when evaluating any backup or DR offering.

By Stephanie Balaouras

Check out Stephanie's research

You should follow me on  here

Deduplication Market Undergoes Rapid Changes

In May, I blogged about NetApp's announced acquisition of deduplication pionneer, Data Domain. The announcement triggered an unsolicted counter-offer from EMC, followed by another counter from NetApp. But after a month of offers, counter-offers and regulatory reviews, EMC ultimately outbid NetApp with an all cash offer of $2.1 billion. I believe that Data Domain would have been a better fit in the current NetApp portfolio; it would have been easier for NetApp to reposition its current VTL as a better fit for large enterprises that still planned to leverage tape. It's also said that more than half of Data Domain's current employees are former NetApp employees so there would have been a clear cultural fit as well.

For $2.1 billion, EMC gets Data Domain's more than 3000 customers and 8000 installs but it also gets a product that in my opinion, overlaps with its current Quantum-based disk libraries, the DL1500 and DL3000. In Forrester inquiries and current consulting engagements, Data Domain is regularly up against the EMC DL1500 and DL3000. EMC will need to quickly explain to customers how it plans to position its new Data Domain offerings with its current DL family, both the Quantum- and Falconstor-based DLs as well as its broader data protection portoflio that includes Networker and Avamar - which also offer deduplication.

In addition to the EMC and NetApp stuggle to own Data Domain, we also saw one backup application vendor after another add deduplication capabilities to their software. The online backup vendors came forward to remind everyone that they have been deduplicating data in software all along. As I completed my current state of the union report on the deduplication market, no sooner had I completed one research interview when another backup vendor would announce new dedupe capabilities. I could have delayed the report forever waiting for a lull in the announcements. I officially interviewed 19 hardware and software vendors for the report but was briefed by an additional 3 vendors in the past 2 weeks. I already know there will be additional vendor announcements in the coming months.

It was also clear as I was writing the report that the definition of many deduplication characteristics are blurring such as inline vs. post-processing and source vs. target. I took hard stances on some of these definitions and forced vendor offerings into one category or another for the sake of clarity but I also tried to explain the nuances of some of the vendor offerings. For IT professionals, at the end of the day, you must understand essential deduplication characteristics and considerations and familiarize yourself with all the vendor offerings in the market but at the end of the day, you must have well defined performance, capacity, scalability, manageability and resiliency requirements. You must issue an RFI and/or RFP, let the vendors make the case for their offerings, short list 2-3 vendors and bring them in for a proof of concept (POC) to validate their claims. Refence customers and POCs are the best way to cut through all the hype, claims, and counter-claims made by the vendors in this market.

By Stephanie Balaouras

Check out Stephanie's research

You should follow me on  here

IBM in talks to acquire Sun?

A big blue cloud overshadowed Sun’s announcement today unveiling their Open Cloud Computing platform. Media was a buzz today at rumors of a possible acquisition of Sun Microsystems by IBM. Still a rumor at this point, the story brings up many questions about how feasible this acquisition really is and if it makes sense from IBM’s perspective as well as Sun’s.

Here is Forrester’s take:

                        First and foremost this would be an acquisition based on market consolidation where a smaller vendor is acquired by a larger player, in an effort to grab a greater total share of a market where growing via margins is impossible. With Suns core business based in servers, a declining market, this seems a perfectly reasonable scenario to envision. The main value for IBM lies in the margin dollars found in the UNIX server market. IBM Global Services is one of Sun’s largest global resellers so it would be an opportunity for IBM to secure this very loyal customer base and increase its share of the UNIX market where all the margin dollars are found in the server market. Add in HP and Sun’s recent partnership around Solaris and you’ve got another big motivator for IBM to look closely at a potential acquisition of Sun to strengthen itself against the competition. Furthermore, IBM has wanted to get its hands on Java for some time as it is a key intellectual property for WebSphere and in terms of Identity & Access Management (IAM), IBM would also dominate the market from a revenue perspective with a Sun acquisition. Forrester believes a combined entity would compromise somewhere near 30% of a $2b+ IAM market. This isn’t the first time IBM has made attempts at acquiring Sun but so far all attempts have been unsuccessful. There also should be no antitrust concerns being speculated as IBM Rational alone has yearly revenue in range with Sun.

                        In terms of Sun, their growth strategy of appealing to developers through open source software — Solaris, MySQL, NetBeans, GlassFish — and hoping they will then turn to Sun for servers and storage won’t work in a global recession. The strategy depended on the rise of a new wave of Internet businesses — Web 2.0 — and expansion of Internet offerings of telcos and existing large service providers called “red shift companies”. In a global recession, neither camp is growing at the rate Sun needs it to. This makes a potential deal a possibility from Sun’s perspective.

                        Still, acquiring Sun would pose a number of integration issues for IBM Software Group. Firstly, IBM aims at executives while Sun aims at executives and beyond with a strong marketing investment towards developers. Secondly, both sides have fundamentally different business models. IBM SWG is not structurally set up to monetize support revenue stream model which they would acquire and very likely would have issues maintaining SLAs required to maintain it. Additionally, on the marketing front, a decentralized marketing model at IBM SWG would find challenges integrating a more centralized model targeted at grass roots like that of Sun. Although these challenges are not insurmountable they do offer a glimpse at several significant deterrents for IBM.

                        While IBM is the vendor which sparked the conversation, other vendors have made bids for Sun in the past and if IBM is interested it’s not out of the question to think others like Cisco and HP might counter. Regardless of the who, if Sun is acquired it should come as no surprise to see the winning bidder to shed a lot of the weaker products (and a significant number of redundant staff) in the process. From a software perspective we would expect Java, Solaris, MySQL, and Indentity Management Suite to live on with the new company with the possibility of GlassFish, NetBeans, Open ESB, and Java CAPS being spun off into open source projects if they are to continue on.
Given the current state of the market any acquisition rumor is food for thought and should be considered but this represents a deal that has just has many pros as it does cons and should be treated as such.

Thoughts on an IBM acquisition of Sun? Let us know what you think?

By James Staten

With Christian Kane, John Rymer, Jeffrey Hammond, and Stefan Reid

Check out James' research

Check out John's research

Check out Jeffrey's research

Check out Stefan's research

What’s New With Windows 7? Not That Much Apparently

Large enterprises and small businesses alike are in the throes of making very strategic decisions about their Windows desktop road map. The result of which is that customer optimism is high for new information on the future of Windows. With Windows 7, however, Microsoft is taking a tighter approach to releasing product information which is driven by Steven Sinofsky, the new head of Windows development. This approach stems from a lot of the lessons Microsoft learned from its Windows Vista experiences and we agree that it’s the right approach to take. Let's face it, Microsoft was burned by Windows Vista for promising too much for too diverse a crowd and it's going to be a little more disciplined about when and how it discloses information on Windows 7. So desktop ops professionals have to be more patient moving forward and discount any speculation that is undoubtedly on the way.

Of course Microsoft is still in the midst of development and testing for Windows 7, so it's way too soon to come to any conclusions. And while the touch capabilities that Ballmer and Gates demonstrated at the Wall Street Journal's "D: All Things Digital" conference last week were neat, they aren't going to help you with this decision-making process. At the end of the day, this news about Windows 7 becomes just another driver to adopt Windows Vista in the foreseeable future. You may be tempted to hang on with Windows XP (especially given the fact that it will still be available on new machines for at least another year), but this will lead to bigger troubles down the road because Windows 7 will be built on the Windows Vista foundation. Forrester recommends that most companies deploy Windows Vista eventually because Windows 7 is clearly going to be an evolutionary update, not the revolutionary update that many are hoping for. So the challenges that organizations are experiencing with Windows Vista today will also occur with what's coming next, and you're better off preparing for the future now than later.

By Benjamin Gray

Check out Benjamin's research

Windows XP Availability Update

With June 30th just one month away, your business may be feeling pressure to upgrade from Windows XP to Windows Vista. But don't panic if you're not ready. Three of the world's largest PC manufacturers have announced that they will continue support for Windows XP on new PCs when customers exercise their Windows Vista downgrade rights. The best part of the downgrade process is that when you're ready to move to Windows Vista, you have already purchased the license for it. So the upgrade should be "free." And why is Microsoft allowing this? Because machines shipping after a Windows Vista downgrade (hence, with Windows XP) still count as shipped Windows Vista licenses, which allows Microsoft to continue touting how well Vista is selling.

Here's a quick guide to the Windows Vista downgrade process by PC manufacturer for all of the businesses that aren't yet ready for Windows Vista.

• Dell: Dell is offering to pre-install Windows XP on newly purchased machines as part of their Windows Vista downgrade service. Dell will provide technical support for the factory-installed Windows XP image. Just make sure you select "Genuine Windows Vista Business BONUS" or "Genuine Windows Vista Ultimate BONUS." There is no additional cost for this service.
• HP: HP is also offering to pre-install Windows XP on new machines until at least July 30, 2009. Just make sure to select "Genuine Windows Vista 32 downgrade to Genuine Microsoft Windows XP." Again, there is no additional cost for this service.
• Lenovo: Lenovo will also support downgrading to Windows XP on many of its machines, but it isn't offering a factory pre-install option. Instead, customers may purchase a Windows XP Recovery CD with the purchase of a Windows Vista Business or Ultimate machine. However, it is only offering this service through January 31, 2009.

By Benjamin Gray

Check out Benjamin's research

Windows XP SP3, The Final Frontier

The last of the service packs for Windows XP (SP3) was released to manufacturing last month and to the Web on May 6 after a delay to fix compatibility with Dynamics Retail Management System (RMS). SP3 contains the usual rollup of all the fixes and patches along with several (mainly security) enhancements including black hole router detection and network access protection (NAP). Don’t expect any new features or new versions of Internet Explorer or Media Player, however, given that Microsoft has turned its focus to Windows Vista and subsequently "Windows 7."

As with any SP release, the Microsoft support forums were immediately flooded with some irate consumers. The most consistent complaint was of endless reboot cycles on AMD-powered machines. This error was quickly traced back to mistakes made on the part of HP that have still not been resolved. However, the majority of consumers and businesses alike should expect a non-event in upgrading.

What does this mean for IT ops professionals? It’s a no-brainer to make this relatively painless upgrade -- especially since independent studies are touting that SP3 adds up to a 10% performance boost in some applications and buys you support from Microsoft through April 2014. Quite honestly, we haven’t heard much from our user clients about SP3. We can’t help but assume it’s because the real focus has shifted to either Windows Vista or Windows 7, but it can also be attributed to the fact that this upgrade is essentially a non-event for most organizations.

By Benjamin Gray

Check out Benjamin’s research

Xiotech Zigs When Other Storage Vendors Zag

Bucking the ‘hardware is a commodity’ trend seen widely in the storage industry these days, emerging vendor Xiotech recently announced their Emprise line of storage arrays based on a modular component called Intelligent Storage Element (ISE), housing drives in sealed DataPacs designed to dampen vibration, prolong lifespan, and reduce the incidence of “no fault errors.”

The ISE relies on a custom enclosure architecture that mounts drives in an opposing fashion to cancel vibration and uses advanced materials and design to improve cooling and IOPS performance. While the drives inside are standard Seagate HDDs, they use specialized firmware to allow for advanced error management and give ISEs the ability to manage around errors, bypassing bad sectors without having to swap out individual failed drives. This has the effect of increasing the size of the base unit of storage, from the individual drive to the ISE, allowing storage admins to spend their time on higher level data management while base level functions like striping, drive, and spindle management are automated.

Technology features aside, the Xiotech solution drives business impact for users. Because of the improved reliability of the new architecture, the vendor offers a five year warranty standard with Emprise arrays, contrasted with typical warranties of three years in the storage industry. This relieves customers from the burden of costly and complex migrations after only three years of service, or the prospect of paying around 15% of total array purchase price for annual maintenance — either of which can add up very quickly in terms of overall IT expenditures. Combine this cost avoidance with the claims for easier management for an ISE based system and the result is a significant reduction of overall operating expense related to storage. Xiotech, a relatively small fish in the ocean of sharks that is the storage industry, gains a serious product differentiator with this announcement — one that can directly translate into reduced deployment and management costs for storage; a timely attribute in this time of shrinking budgets and ever growing data.

By Andrew Reichman

Check out Andrew’s research