Thursday, September 17, 2009

Oracle-Sun Hardware: Easy to say and Hard to do – Oracle's Exadata 2

More than 10 years ago I was invited by Oracle Distributor or Oracle Israel (I do not remember if an Oracle's local branch was already established) to a dramatic announcement by Larry Ellison. Tel-Aviv was one of the cities in which you can watch it in real time. Oracle in cooperation with Sun announced the age of Network Computing.

It was a good show. Larry Ellison broke a PC demonstrating that Microsoft's Fat Client applications and data were lost. He also broke a cheaper Network Computer (NC) and demonstrated that nothing was lost after using another NC for connecting to the Web. Network Computing was immature. It took some years until Google's applications as well as other Web 2.0

Services vendors' products were used and more than 10 years until SaaS is going

Mainstream and Cloud Computing services are used for some enterprise applications. Oracle and Sun anti-Microsoft campaign continued with J2EE vs. .Net wars. However, few years afterwards both companies signed cooperation agreements with Microsoft.

I recall this brief Oracle's history while reading articles on Oracle's Exadata version 2. According to some of the articles "He (Larry Ellison)'s so furious that he's dumping HP as the hardware partner".

It is true that it took a lot of time until USA regulators approved Oracle's Sun acquisition and that the deal is waiting for European regulators approval, because of database and Java anti-trust issues.

It is also true that IBM and HP and to lesser extent Dell, take advantage of the delays and snatched some of Sun's hardware customer Install Base.

But I would doubt if that was the reason for Oracle to abandon the Exadta partnership with HP described by Ellison as "Oracle's most successful introduction ever" few months ago. As the Oracle-Microsoft history cited above shows Oracle's decisions are not emotional decisions based on Ellison's Love, Hate or Anger, but rather Business centered decisions.

As I already mentioned in previous post Oracle is not a hardware vendor and in my opinion (which is supported by other hardware vendors snatching Sun's hardware customers) supporting and extending Sun's hardware Business Lines is a major issue.

I was also skeptic about Oracle's Vision of Car like combined Hardware Software products.

I will analyze Exadata 2 technical aspects in a next post, but it should be noticed that although it is based on Sun's hardware solutions it is based on standard Intel processors and not on Sun's SPARC processors, supporting the view that SPARC processors are nearing their End of Life period.

Another concern for Oracle and its customers is: What should current users of Exadata 1 customers do?

My Take

I still do not think that Oracle is a Hardware vendor and may not be in the future.

Exadata 2 announcement does not prove the thesis that it will become a hardware vendor.

I am not sure that other storage vendors together with Database producer will follow Oracle (IBM, HP with Microsoft, EMC with Microsoft, Netapp with Microsoft).

It is too early to reject the hypothesis that Oracle's Sun acquisition deal will turn into selling Sun's hardware Business Lines to HP or to another hardware vendor or vendors as described by some analysts after the Sun's acquisition announcement, raised again in IT Web journals few days ago and mentioned as a possibility in one of my Vendors Survival posts.

Oracle may sell only viable part of Sun's hardware business lines.

If Exadata 2 will be successful it may be included in such a future deal. .


Meni Meller (מני מלר) said...

Its true that Oracle holds all the cards very close to its chest. Its true that uncertainty is here.

Its also true that every Oracle customer has year to year growing storage needs which EMC, NetApp or now Oracle can fulfill.

I think that when Oracle R&D guys sit with Sun's HW R&D guys many innovative solutions that simply can't exist today will emerge.

At the bottom line, I think that if Oracle (which is a SW only company as of today) would like to significantly grow their business, they need to do something creative, innovative and most importantly DIFFERENT than what they / their competition can / does in its day-to-day reality.

I don't think its right for Oracle to start selling pizza box servers and i'm sure Oracle knows that. The edge, the money -- the ROI can only be achieved by creating a fresh new solution, something of the kind that will surprise our industry.

Thinking about it more in-depth, its just like owning Java.
Java is virtually the hardware on top of all Oracle's SW runs on (except for the DB which is C++ IIRC), owning Java helps you innovate your currently existing Java applications / Servers. Owning the HW design will help you do even more.

I hope that his is where Oracle is taking Sun's technology for, I could always be mistaken, but it just doesnt feel right, and raises a lot of other questions, financial, regulatory and simply logical...

Avi Rosenthal said...

meni, thanks for your comment.
I do think that Sun was an innovate and creative company (e.g. creation of Java), which earned my sympathy.
Oracle is less creative and more business oriented. If you are a small company (like Google in its first years) you can be very creative. If you are a very large company like Oracle or Microsoft, you have to address your customers, requirements. The most creative Database makers in the 90ths was Informix. Oracle is still the DBMS market leader and Informix is no longer a vendor.Large companies like Oracle usually buy inventive companies instead of being very creative (sometimes Oracle and other large companies are creative). You can read a relevant post in my blog titled Vendors Survival Guide: Supermarket, Grocery and Kiosk .
The bottom line in my opinion: Oracle is not a hardware company and if it think of combined Hardware & Software product it could easily partner with an Hardware vendor. Successful Exadata 2 or 3 could be sold together with other storage products to a storage hardware vendor

Public Cloud Core Banking: Hype or Reality? - Revisited

  More than 4 years ago I was asked if Public Cloud Core Banking is a Hype or a Short Term Reality? If you had read the post, you would prob...