The turn of the century was a bad omen for computing industry stalwart Sun Microsystems. Once seen as leading the pack in the Unix server market, its decline in fortune may have stemmed from its refusal to accept early on the potential of the Intel x86 platform as a viable, computing workhorse for all but the most demanding of applications.

As far back as 2002, Sun faced quarter-upon-quarter of revenue decline even as analysts estimate the market was actually picking up steam with rising server revenues by most vendors (except Sun).  And so it was that nearly 8 years later, Sun finally conceded defeat accepting an offer by database giant, Oracle, to be acquired for US$7.4 billion at 2009 levels.

Should the industry mourn the death of Sun? I don’t quite revenue the atmosphere during the days when Digital Equipment closed it’s doors following the completion of its acquisition by Compaq (itself eventually gobbled up by HP).

Will the sunset for a veteran hardware vendor mark the official beginnings of the dawn of the software age? It can be argued that for many years most industry observers swooned to the music of hardware vendors. Even today we applaud with each new processor by Intel, or the new lineup of Thinkpad laptops, and tomorrow – January 27, 2010 – the much anticipated Apple tablet device (or whatchamacallit). Sure, we turn our heads when Microsoft launched the latest incarnation of its much despised yet very popular Windows operating system. Yes, enterprises raised their hands to view the latest SAP ERP software. And definitely, businesses are listening more intently on how Software-as-a-Service will reduce their CAPEX cost considerably and make them look better on the accounting books because OPEX doesn’t hurt their market positioning as much as CAPEX.

Its hard to figure out when the software revolution started. But you can bet that just as Apple revolutionized the MP3 market not with a neat, flashy, fancy music player (on the contrary it defined convention by being overly simple) but with software, so too will we finally see the years ahead as the period when software defined how consumers and enterprises will use technology.

For the moment, we bid fond adieu to one of the pioneers of hardware-based computing solutions – Mr Scott McNealy. He is, by many reckoning one of the more colorful characters of Silicon Valley. Hopefully his legacy will somehow survive under the watch of Oracle CEO, Mr Larry Ellison – another industry stalwart.

CEO farewells are fun to read because they are often drafted by wordsmiths who don’t fully understand the emotional turmoil that accompanies an executive’s departure. I am not sure if Mr McNealy hired a professional writer for his farewell but it certainly paints a sad story of the rise and fall of an icon. So before you take out that tissue to wipe away the sadness in his farewell message, watch this video to take the bite out of Mr McNealy’s bittersweet farewell.

Click more for the memo. (more…) recently posted its commentary following IDC’s 2010 prediction that IBM would acquire Juniper Network. This may be a response to HP’s ongoing acquisition of 3Com and the continuing speculation of a potential Cisco-EMC merger.

The much delayed completion of the acquisiton of Sun Microsystems by Oracle will likely come to conclusion during the first of 2010.

Seeing itself more a bridesmaid than a bride is Brocade, which has put up its “For Sale” sign and seen no real suitor moving to take the offer. As far back as October 2009, Gigaom analysts pointed to Juniper as a potential buyer for Brocade. IBM would have also been a good suitor but the overlap is in the Foundry products (assuming Big Blue does acquire Juniper).

I still think Brocade would made a fine addition to HP. Of course Brocade may need to get rid of its Foundry portfolio to minimize overlap with HP’s Procurve offerring.

Assuming that Cisco and EMC do combine, that would leave NetApp and HDS as the lone pure play storage vendors. Can these two survive in an environment where customers want true ease of use, best possible integration, and one throat-to-choke accountability from their vendors?

If you look back at all this a proliferation of vendors came about because customers clamored for more choice. But while enterprises say they like choice because it allows them to haggle on the price, in reality, they will turn to one vendor who has everything because its simply easier to deal with one supplier than a multitude of vendors.

Having a choice of sources is great only in as much as it gives you the perception of freedom to pick what you want. But choice brings chaos to the equation. This goes agaiinst the grain of businesses that must operate in an orderly fashion so that processes can be streamlined resulting in greater efficiency, higher productivity and lower total cost of doing business. Which if my math is correct will mean greater profits and better shareholder value.

So the old adage of vendor lock-in doesn’t hold sway in today’s highly competitive world. As one CEO told me recently, “At the end of the day, its really not a matter of giving customers a wide array of choices to choose from. It is giving each customer the product that he or she will want to buy. And that could be just one product it just so happens to look a little different to the individual customer.”

Smoke and mirrors!

October 10, 2009. I braved the cold winds of San Francisco to attend Oracle Open World 2009. It was my first time to attend this event and true to what my hostess told me, it was jam packed with people wanting to listen to Larry Ellison deliver his welcome keynote address. What surprised me was Scott McNealy, chief executive of Sun Microsystems peddling his stuff to a crowd of Oracle guests and employees. Everyone was expecting McNealy to give his traditional Top 10 “funny, sometimes insightful, and almost always highly Microsoft-targeted” list. But the on-going Oracle acquisition of Sun prevented him from being his usual self (or so he said). Watch him deliver what may be one of his last keynotes as CEO of Sun Microsystems.

In his keynote, McNealy reminded everyone that Sun has “always been about innovation and despite today’s technology having the shelf-life of a banana, you have to keep inventing, creating and breaking-through with new stuff.” He also emphasized that the combined Oracle and Sun companies will create one of the largest pools of R&D in the world. Innovation is not just about creating something new but you need to provide value to your customers, and must result in success.

Larry Ellison, for his part, shared the ‘same excitement’ about the potential that two industry innovators (Sun on the hardware and Oracle on the application software) can bring together to the market. He promised to continue investing in Sun’s core innovations including the Sparc chip and the Solaris operating system.

The announcement of the next release of the Exadata server series seems to corroborate this promise. The first Exadata released was a HP-Oracle venture. Exadata Version 2 sees Oracle partnering with Sun to produce an even better product. While the first version focused on speeding up data warehousing (DW), the succeeding version included online transaction processing (OLTP) compute capability in the same box. What Oracle is trying to do here is deliver a complete DW/OLTP solution that integrates server, storage, operating system, middleware and application to do something faster than is possible by amalgamating different elements from different vendors, the way system integrators do.

Certainly the Exadata offers a slightly radical way of running your data warehouse. If you believe Larry Ellison’s slide on the comparative cost of the Exadata to the IBM near equivalent offering, the price differential is staggering. Such a shift in the way data warehouses are built and delivered should amount to what McNealy calls innovation that delivers true value and profits to an organization.

So am I sold on the idea that Sun will thrive under the Oracle umbrella? The Exadata 2 proves there may be a future for Sun inside Oracle. But I still have some lingering doubts and perhaps its because of the way McNealy left the stage as he closed his keynote. In my mind, it signalled a quiet resignation on the part of this entrepreneur who co-founded a company that today represents a sizeable install base in finance, education, and government industry sectors. But maybe I am just reading too much into it. Watch for yourself below.

Click here for other stuff around Oracle Open World 2009.

The possible IBM acquisition of Sun has finally settled. Software giant, Oracle, has come out holding the Sun trophy.

Hang on, where did that come from? Most everyone in the media appeared to have been caught off-guard. Why would Oracle, the world’s 6th largest software vendor according to Software Magazine even consider such a deal? What could Oracle possibly gain from a Sun acquisition?

Oracle CEO, Larry Ellison, seems to think this is a good fit. “Oracle will be the only company that can engineer an integrated system – applications to disk – where all the pieces fit and work together so customers do not have to do it themselves. Our customers benefit as their systems integration costs go down while system performance, reliability and security go up.” was posted on the Oracle press release issued April 20.

The same carefully drafted press release also appears on Sun’s website and quotes Sun Founder and Chairman, Scott McNealy as syaing: “Oracle and Sun have been industry pioneers and close partners for more than 20 years. This combination is a natural evolution of our relationship and will be an industry-defining event.”

The deal would suddenly propel Oracle into the Tier 1 server vendor marketplace competing head-to-head with IBM and HP for a place in the analyst polls of leaders. the acquisition also gives Oracle a leg up in the storage where Sun has a dominant position in enterprise tape library business, thanks to its earlier acquisition of StorageTek, and a stake in the promising open software storage space.

Where are the overlaps? According to an IDC 2008 report, Oracle led the worldwide database market share in terms of revenue growth with 44.3% far ahead of archrival IBM’s 21% share. What happens now with MySQL – arguably one of the most popular open source database software in the market today?  Will Oracle throw it out the same way it did to Innodb when it acquired Innobase in 2005?

Some argue that we will not see a repeat of history. MySQL does not compete with Oracle 11g. There are literally world’s apar: the former is ‘almost’ free whereas Oracle’s flagship product makes a significant dent in most IT budgets – if you could licensing and software.

There is an interesting blog discussion on reddit about the potential for Oracle to offer MySQL as a lite version of Oracle database. But as with all other blog discussions I’ve read so far, people are thinking MySQL is not a threat to Oracle’s database business. Rather, the service and support side of MySQL is something that would be a plus to have given the very large community of MySQL users in the world. How to make them all pay for support without causing an exodus to other open source alternatives will be the challenge.

Another big bonus to Oracle is Sun’s large install base of happy customers in government, telecommunications and finance. This is something IBM sales folks will be salivating over, and fearing, as now Oracle sales people start cozing up to the same customers for database upgrades (DB2 out! 11g in!).

The really big question is what to do with all that hardware? Oracle is not a hardware company. The one time it did try to become a hardware vendor was when it launched the Network Computer, a collaboration with Sun. Despite much media coverage, that product never took off. 

 If the vendor is to ever succeed in the hardware business, Oracle will have to work hard to make sure it keeps the right Sun talent on board especially for the duration of the integration, and it wouldn’t harm, if they stayed on even with an Oracle badge and business card. Expect to see a long, hard road. Easy enough to predict an exodus of talent from the original Sun camp.

Theresa Lanowitz, founder of Voke Inc, raises the potential clash of cultures – Oracle is a software vendor whereas Sun has hardware hardwired into its DNA. “Sun does not understand how to develop, productize, and market compelling software,” says Lanowitz. She also predicts dire future for a number of areas of practice within Sun in an Oracle future (or rather lack of future in an Oracle world).

What’s next for Oracle? The vendor has shown a strong knack for acquiring software vendors. Perhaps this is where it should keep its focus. What about Red Hat?

Finally, a quick recap of what IBM lost in the deal. It lost JAVA to which Big Blue has made public its commitment to the platform. Oracle gets instant access to the very large MySQL database community, and the potential service revenue that comes with it. With the Sun Solaris tucked neatly under its belt, Oracle has the opportunity to “finely tune Oracle database for some of the unique high-end features of Solaris.” On the hardware side, IBM would have had 100% market share of enterprise tape library business following a Sun deal.

All this for a measely extra US$1Billion? Oracle bought Sun for US$7.4 Billion compared to IBM’s offer of US$6.5 Billion.

The April 14 announcement of the new EMC V-Max storage array raises what I have long-held is the reckless abuse of technology terms to win over customer (and media) attention. In the press release, I note that ‘virtual’ appeared 36 times. I am trying to remember what my English teacher in elementary grade told me about the repetitive use of words. I guess not enough marketers are being shot for single-minded ignorance of English grammar rules. Anyway back to the story at hand.

Practically every major storage, server and system vendor announcement in the last 18 months have “virtualization” as a central message. Everyone is riding on the bandwagon of “do more with less”. The technical promise is optimization or better utilization. The financial promise is less money.

The problem is too much marketing hype is creating confusion.

The technical definition of virtualization is the abstraction of compute resources. Wikipedia lists six different types of virtualization depending on what the “resource” is. The layman’s definition of virtualization doesn’t really exists. It depends on what the base technology is. If you are saying, it’s about the operating system, then virtualization means running many operating systems (for example Windows) on a single physical hardware. If you think applications, it is running many different applications on the same OS. If it’s data storage, it is having access to data no matter where it is physically stored.

Confused? I know I am!

Anyway, EMC – arguably the dominant data storage vendor today (for now) – launched their latest storage array on April 14 (US time) with virtualization almost literally as the adjacent label to the product. The new Symmetrix V-Max is a humongous array that promises pooling, migration, management, functinality and asset re-use all from one single product. (more…)

This morning I woke up to Bloomberg reporter, Ian King, explaining to Bernie Lo, Bloomberg’s Hong Kong anchor, that IBM is talking to Sun [8] about buying the latter. King pointed to a story on the Wall Street Journal [9], citing the acquisition of Sun by IBM [10] would bolster Big Blue’s “heft on the Internet, in software and in finance and telecommunications markets.”

What a lame excuse! IBM has a significant presence on the Web, just read Lou Gerstner’s autobiography, Who Says Elephants Can’t Dance [11]. Most people may not know it but IBM is a software company too. Its been in the Software Magazine 500 Software companies list for several years. Its mainframe-related software business alone should dwarf many pure play software vendors out there. Granted it doesn’t have JAVA, a Sun original IP, but it has its own string of innovations from time immemorial. MySQL is another story though.

Anyway, this is a storage blog so let’s stay focus. Relative to IBM, Sun’s storage business is mostly OEMed from other vendors – LSI [12], Engenio [13] (acquired by LSI), and Dot Hill [14] (lowend to midrange), and Hitachi (enterprise). Its tape is largely an acquisition – StorageTek. So from a hardware perspective, Sun doesn’t have much of a home-grown offering. But who cares? Certainly not Sun customers. They buy Sun for its solution sets, its expertise, and its history in high performance computing workstations and depth of knowledge/experience of the Web. Oh yeah, its Open Software strategy has been at the forefront of media coverage too.

So why would IBM buy Sun? Beats that pants out of me. IBM has a full range of storage (disk, tape and software) offerings – some of it OEMed, some of it as a result of acquisitions, and some of it OEMed. Its 2007-2008 acquisition list [15] is impressive. Diligent Technologies (data deduplication), FilesX (continuous data protection), XIV (SAN and clustering), Arsenal Digital Solutions (data protection), NovusCG (storage management of heterohenous environments), Princeton Softech (data archiving, data classification, data discovery), and Softek Storage (data migration).

So why buy Sun?

Acquisitions usually happen to (1) bring in new technology that the buyer doesn’t have; (2) take out a competitor and grab marketshare; or (3) get the customer base. In this case, I’d speculate it may be a combination of all three but certainly individually, Sun doesn’t have much to offer IBM. Sun’s Open Source Storage Strategy has a long-term potential to throw a spanner on IBM’s storage management software business. Heck it has the potential to mess around with everyone major storage vendor’s software strategy today. ZFS promises to make data software management ‘almost free’.

In the IT enduser community, the idea of Open Source brings with it a plethora of elation as well as fear. Ask anyone who is an IBM AS/400 and they will tell you that they are stuck in their for life – at least until they decide to get out of what they put themselves into.

Yes, IBM AS/400 (rebranded as IBM System i in 2006 and subsequently replaced by the IBM Power System line) is a very stable platform. The many applications developed for it are rock solid, enterprise-class software that do what they are meant to do. Throughout its period of reincarnation (1988 to present), the hardware and software may have changed but IBM made sure the applications are transplanted. 

I am digressing so let me paraphrase one CIO comment about their AS/400 investment. “We are stuck and we know we are paying through the nose but we have no alternative today!”

Distributed Computing (DC) arose partly in response to the need to get out of the mainframe and mini-computer era of vendor lock-in. Little did we know that while DC hailed the arrival of an army of vendors offering competing and complementary systems, the liberation was partial – because many of the initial technologies created in support of DC are proprietary in nature. Sure they are able to talk to other vendor’s solutions but this is because application programming interface (API) were built to allow for some semblance of interoperability.

Enter Open Source. The idea that a program’s code is freely available to the end-user community to use and modify to suite a particular need. Can a software company survive giving away its software? Red Hat thinks so. In fact, within the Open Source community, Red Hat is a testament to the idea that you can give away copies of your software (even if it was originally conceived by someone else), make money and prosper.

SUN Microsystems is another company that is heavy into the Open Source momentum. But whereas RedHat is 100% open source, SUN still has technology that is proprietary – afterall, SUN started life in the proprietary world. It can be argued that JAVA was SUN’s first experiment in Open Source. Thankfully, the JAVA community thrives today despite competition.

As with all things, proven or otherwise, there are skeptics. In human nature, the biggest fear is always that of the unknown. For many enterprises whose businesses depend on the smooth running of mission-critical applications, the high price associated with proprietary hardware, middleware or operating systems, custom application, and availability of skilled resources is a bitter sweet pill that they’d readily swallow.