Rod Adkins, the SVP and Group Executive of IBM’s Systems and Technology Group (STG) took the time to engage the influencer community quite early in his tenure for a well-run event at the Watson Research Lab in Yorktown Heights. “I’ve been in this position for 38 days,” he  reminded us, as STG’s AR team widened the usually hardware-focused invited audience to include generalists and more software-focused folk like me.  IBM execs from IBM’s Software Group, its Research organization and corporate, joined us  for a look at the science behind the systems, a compelling addition to the agenda. And another pitch for IBM’s analytics thrust was a scene-stealer.

Adkins is hardly new to the party: he spent 10 of the last 12 years in STG in addition to the work on pervasive computing he did in the Software Group. “We play a vital role in the Smarter Planet initiative – we have to provide optimized infrastructure to capture, manage and deliver all the data it runs on,” he pointed out. He clearly articulated a strategic path for 2010 to match his aggressive influencer communications agenda: growth in workload optimized platforms, systems software value capture, delivery model changes and data center wins. For each, organizational changes, investments, messages and new offerings are well scoped and already underway. STG has 47% of IBM’s R&D budget and is investing in process and packaging; technology design; hardware systems; systems software; and client support. Yes, research into client support – a topic well worth its own discussion, but beyond our scope for this post.

Adkins takes the helm of a ship in good order: 2009 performance to date has been steady, and compares well with its peers given the economic environment. Revenues for Q3 were down 12 percent from the third quarter of 2008 (better than HP’s 17% decline in its server group). IBM’s mainframe, late in the z10 product cycle, decreased 26 percent. It’s not unusual to see declines at this stage as the market anticipates the next (usually well known) version. Still, z continues to defy all predictions of its demise: it has almost doubled its share of systems over $250K in the last 8 years from 17.2% (Q400) to 32.1%(Q408). Adkins touted IDC and Gartner reports of market share gains for Power Systems, System x and disk and tape storage during the third quarter. According to industry analyst sources (IDC and Gartner) revenues from System x servers increased 1 percent; microelectronics OEM revenues were fairly flat at a 1 percent decline. The 45 nanometer business is running ahead of the previous (65 nm) ramp – its production run is already sold out and fab performance is good. IBM’s acquired XIV storage has shipped its 1000^th unit – 70% of which went to new customers. And IBM claims to have capitalized very effectively on the turmoil in Sun’s base: “We’ve won business from 100 of Sun’s top 300 customers.”  Even system software has a few things to crow about: the Migration Factory solution, IBM asserts, has driven 2000 migrations to IBM POWER systems so far this year. 

IBM is very optimistic about upcoming technology introductions: its next generation Power7 chip architecture will feature substantial performance and energy consumption improvements, as well as some interesting new dynamic thread management capabilities. Big bets are being placed around the notion of “workload-optimized systems,” pre-integrated solutions such as:

• Cloudburst  private cloud hardware appliances for secure deployment and management of  application environments
• The new scale-out NAS (SONAS), for standards-based, large scale file storage
• Smart Analytics System,  a pre-integrated  analytic appliance that delivers IBM data warehouse software preinstalled on an IBM server and storage

Some innovations are delivered as what IBM is calling “system optimizers.” For example, the Smart Analytics brand is extended to the Smart Analytics Optimizer, a software-based accelerator (available now for z and moving to other platforms) that automates the movement of frequently queried data into as much as a terabyte of main memory, where other tricks like vector processing and a new “optimized storage format” (I haven’t explored edetails of his yet) IBM claims will dramatically improve performance. IBM’s decades of database  optimizer research enable it to drop this in and “know” when to push queries to the in-memory database. This kind of hybrid system will change the architecture of data management if it is proved to work as it rolls out – and I wouldn’t bet against it.

IBM has been ramping its geographic coverage with 56 added global branch offices and will add 15 more in 2010. The channel is getting more investment in education, enablement, and training in how to improve profitability. And STG has created a new System Software business unit to drive development, revenue and margin growth, with a service management focus that will take HP on in one of its core areas of strength.

Where STG Fits in the Smarter Planet Story

John Iwata, IBM’s Senior Vice President Marketing Communications, gave us a conversational look at how IBM continues to apply deep probes into its own business to understand company customers, their needs, and resulting opportunities. “We looked at 20 showcase IBM solutions that we had built and implemented for clients,” he said. “We found that all 20 had the entire portfolio of IBM product and service, technology and in almost all of them the Research division played an important role in the eyes of the client as differentiation as to why they chose IBM, and” – perhaps most interesting – “are  instrumenting their physical systems. By 2010 there will be 1B transistors per human; there has been a hockey stick in deployment. The internet of smart things, and the associated data, are the wellspring of IBM’s Smarter Planet thinking.”

By now, some of the examples are familiar but Iwata unveiled one I hadn’t seen before: smart sewers. Who would have thought? But IBM predicts at least $70B of expected investment in sewer modernization. In San Francisco, smart manholes have a float attached to a wireless transmitter to gauge water flow and water level.  Our sewer systems, many of which are a century old, are built for average volume, but when you have above average flow they are easily overwhelmed. It’s a familiar phenomenon. And worse things happen: as I write this, a 100-year old main is being repaired after rupturing two days ago and causing massive problems, creating a giant sinkhole flooding streets and buildings, and closing street. Perhaps future systems will be able to anticipate these failures better.  Iwata pointed out that other instrumentation might let you learn that a certain pump needs to be fixed; doing so beforehand avoids similar failures and catastrophes. And, of course, cities will get smarter than “managing by averages” permits by tracking in real time, and learning from the data.

Iwata noted that “We don’t label anything ‘dumb’ in our messaging. Nobody designed the systems that way.” Advances in technology permit new ways of thinking, even though, as he acknowledged, the timing could not have been worse in terms of the economic conditions. It’s a leadership agenda, not just an ad campaign. IBM took Smarter Planet to governments all over the world – it can raise job growth. Much economic stimulus money flows here.  Iwata has been delighted to see how the field has taken this program up enthusiastically – they don’t always, do so, he acknowledged wryly. Over 80 field-driven events have been done, with Ho Chi Minh City being one of the most recent, focusing on a “smart food” supply initiative. Emerging markets are looking to leapfrog from traditional systems to post-industrial, information age ones. IM stands to gain enormous opportunities from all this. It can do well by doing good.

The Mathematician’s View of Advanced Analytics

The star turn from the wings at this event came from Brenda Dietrich, a last minute substitute when a snowstorm in Minnesota prevented a Mayo Clinic presenter from joining the meeting. Dietrich, an IBM Fellow, and VP from Research, has made many contributions over an illustrious career; she shared with us that she coded the 1959 algorithm underlying GPS as a new intern and was the largest user of computing power at IBM that year. Today’s GPS device makes for a good example of the potential for advanced analytics, when one thinks about it.

Dietrich says analytics can be applied in many places – for example, she and her team have been modeling IBM’s biggest customers, and their own sales organization – and have made recommendations on making adjustments to the configuration of their sales force. Some of the changes were implemented – and they’ve seen large lift (over $1B.)

Dietrich pointed out that the state of the art in statistics is to “extract the explainable and extend.” Of course, sometimes that’s not as easy as it may sound. For example, the Super Bowl occurs on a different day every year – how to put information on its impact on your sales back in to your predictive model to leverage it in retail (to sell more big TVs, or just chips and beer) is not obvious. She also talked about the frontier cases – such as streaming data, which opens the door for a new type of statistics. Massive data sets from the mobile web, like GPS and OnStar, could be the basis for exciting new opportunities if used effectively. And her proximity to the STG teams gives her a view of how the science of analytics can benefit from new architectures. She noted that “most of the algorithms in the field assume serial, not parallel, hardware – so a lot of work on things like matrix inversion can be rethought.”

Finally, as a truly provocative thought, Dietrich challenged us to consider what would happen if computers got smart enough with type checking at a higher order of computation. What if a built-in help system could tackle the garbage in, garbage out problem by telling users “I’m sorry, but there isn’t enough data to make that prediction?” Think of it as “warning labels on mathematical models.” Might have been nice for the quants on Wall Street to have had that – although there’s no guarantee they wouldn’t have ignored it, as they did with other warnings that Dietrich said ware not only visible, but had been noted.

It was easy to walk away from this presentation impressed by the inside look at how advance and predictive analytics are thought about – and many of the attendees were. Even more intriguing was the obvious connection to Smarter Planet – it’s in the very nature of these techniques that they can do things with the newly instrumented physical systems that have not been thought of before. Delivering on the promise is one thing, but IBM is after something bigger: finding opportunities that have not been thought of yet. Few would have thought of mathematical work of this kind as creative endeavor, but Dietrich’s enthusiasm, command and passion make it clear that it is just that, and the best is likely yet to come.

Oracle has been making much of its recent benchmark results. Its new TPC campaign may backfire, however; its deceptive assertions do it no credit, and obscure some interesting technical advances (such as its first use of flash technology) behind mislabeling and deliberate omission of important facts. The “benchmark wars” are far less active than they were in their heyday, when new leapfrogging results occurred quarterly, or even more often. TPC-C, the transaction processing measure, has long been understood to be a poor representation of today’s real transaction types. At various times, most of the DBMS vendors have stopped issuing them – but they come back when they think they can get a headline or two. Some hardware vendors have also been dismissive of the benchmark; in fact, until this one, Sun had been a skeptic for a number of years.

In practice, most production transaction processing requires DBMS features which are routinely turned off to achieve the breakthrough numbers that vendors like to tout. TPC-E was developed to correct some of these issues, but a quick glance at the top results shows that only Microsoft SQL Server results are available so far. Oracle and IBM DB2 have stayed with TPC-C. 

Those busy benchmark days a few years ago were driven by competing hardware architectures as much as database providers, and it’s no surprise that recent interesting results are being driven by hardware again. Kickfire has grabbed attention with its hardware-based SQL processing to deliver some extraordinary results on TPC-H benchmarks, and column stores running in memory with “smart storage” are likely to roil the waters there some more in 2010. But the most visible TPC benchmark in Q4 was Oracle’s TPC-C, and not surprisingly, it was driven by the desire to tout a hardware play: Exadata v2. Unfortunately, the benchmark Oracle published is not an Exadata benchmark at all – not even V1 – and some of the interesting things about it are obscured by the manner of its description, unless you dig a little.

A quick glance at the published Top Ten TPC-C results as of December 13, 2009 (see http://www.tpc.org/tpcc/results/tpcc_perf_results.asp) shows that Oracle has indeed delivered the top recent result. “Recent” is important here: the report was submitted November 3, 2009, in time for Oracle Open World, but after the first ads ran. The report was preceded by a prime media buy (Wall Street Journal and elsewhere) in October that compared it to IBM’s DB2 benchmark from 15 months earlier and promised to beat it.

How that comparison was done publicly is instructive, both for what it says about the performance of the product, and for the manner in which Oracle promoted it. We’ve seen this style before; it can be characterized as at best overstated, and at worst deceptive. It’s based on the premise that the headline is what people remember – few people would read the fine print details if there were any. 

But in fact, there wasn’t any fine print on the benchmark details in the advertising, and that takes its meaning into different territory. After paying the (very unusual) fine against Oracle imposed by the TPC, which explicitly forbids comparing unaudited, unsubmitted benchmarks to posted ones, Oracle subsequently issued ads which didn’t actually contain the tpmC or price/tpmC numbers – and weren’t about the Exadata machine at all. The fine was trivial – amounting to far less than any of the ads themselves cost. One is left to assume that it was just considered part of the cost of the campaign: as a founding and very active member of the Council, Oracle is certainly not unaware of the rules.

Digging a bit deeper (I’m not given to microanalysis of full reports anymore, but a few things jump out from the Executive Summary, compared to the one for IBM’s #2 result - both are only 4 pages), the actual results demonstrate several interesting things about the state of the art in transaction processing:

• Clustered systems have made great strides in delivering results. This is the only clustered system on the list, run on 12 systems; all the other results in the list are on non-clustered systems, and they are all behind in absolute numbers. Clustered architectures are now demonstrably capable of competing at the top end (if there were any doubt.).
• Clustered systems deliver a lot of capacity/power for less. Oracle’s benchmark ran on a Sun server with 48 processors, 384 cores and 3072 threads, a configuration with a total cost (including storage and software; more on that below) of $18M in late 2009. IBM’s had 32 processors and 64 cores, 128 threads, but it cost $17M in mid 2008. So from the system cost perspective, a little more money (relatively speaking) now buys more than an order of magnitude (24x) more threads. Again, perhaps not surprising, but one nice thing about the TPC is that it represents a stake in the ground; theoretically, anyone can buy this configuration at that price.
• The results fall far short of the hardware uplift. Though the news is good on the hardware side, the tpmC uplift: 7,646,486 for Oracle vs 6,085,166 for DB2, or 25%, seems almost trivial by comparison. Oracle’s statement that “Oracle and Sun were able to set the world record using eight times less hardware than IBM used for its largest benchmark” was one of the questionable choices of phrasing in their messaging about the results.
• DB2 clearly wins the “per core” comparison. Looked at on a tpmC per-core basis, Oracle got 20K/core, while IBM got 95K/core. I’d be careful about per-core licensing under such circumstances; the good news is that Oracle recently reduced their core multiplier, clearly seeing the value in pushing customers onto machines with more cores. (No doubt it helped the price/performance comparison for the benchmark too.)
• Memory drives performance, but not as much as one would hope.) There’s a sizable difference in memory: 6 Tb for Sun (at 512 Gb per node) and 4 Tb for IBM (64 Gb per core). A 50% difference, but not a 50% performance uplift. As system architectures become more effective at using multiple cores with dedicated memory, we should see a series of jumps in performance ahead.
• Solid state storage is transforming economics, if not yet performance: the Sun system used 686.6 Tb (with a good deal of new, high-performing solid state disk), which cost $8.4M; IBM, using older disk technology, used 746.5 Gb. at $11.5M. This is an extraordinary volume difference, and it suggests the economics of storage are shifting as rapidly as those of processors.  But again, all the added volume does not seem to have driven a proportional improvement in performance.

Bottom line: this suggests there is a very interesting set of skirmishes ahead, and substantial improvements in results I’d expect to come from both vendors. When IBM delivers benchmarks on its current DB2 release (the one discussed here was on DB2 9.5, not the current 9.7), current hardware, current storage – and perhaps using its own clustering pureScale technology, the game will really be afoot. One hopes IBM’s discussion of the results will be somewhat less fuzzy than Oracle’s. In the meantime, they can only stand by and watch the big headlines sink in; IBM rarely gets into you said/we said kinds of games. At its recent conferences for analysts, there was no talk of these issues to match the red meat attacks heard at OOW. They need to get their own benchmark to market. Soon. Watch also for a wildcard if VoltDB, the stealth project now in early trials, decides to publish an audited TPC-C.

The economic slowdown was not kind to SAP in 2009, and as it launched the annual Influencer Summit on December 8th, change was in the air. Messages were shifting. ”Sustainability” got a big push, and there was a ringing commitment to substantial, dramatic product change to be delivered in 2010. Different faces were on display: there was no Leo Apotheker or Bill McDermott on the stage, although Board members Jim Hagemann Snabe and John Schwarz held down the fort with new Marketing EVP Jonathan Becher and CTO Vishal Sikka in key speaking slots. Like the dances I went to in high school, the event was mostly date-free, but direct questions elicited some specific, though uncommitted, statements about deliveries in 2010, especially from Marge Breya.

 Becher spent too much time on silly stage business interacting with a “talking computer,” but made some important points about changes in technology architecture that are driving adaptive responses within SAP’s product portfolio. There was much to talk about. But 50% of its customers are not on the current ERP release (SOA-enabled and thus requisite for new features) after 6 years. The Business ByDesign midmarket play (one of 3) is still in beta, ramping at a snail’s pace (Snabe said Business ByDesign has “close to 100 customers.”) So there is much work to do. Although SAP asserts that 75% of its 92,000 customers are “SMEs” by its definition, 45,000 of them are Business Objects customers.

SAP faces significant economic challenges. It’s already well penetrated among enterprise firms, and fewer greenfield opportunities remain at the top. New giant contracts will not be in plentiful supply; SAP must find a volume model that works. Despite its continuing addition of new products targeted to the same base, SAP must grow in the midsize market, and that’s not going well while competitors like NetSuite and salesforce.com – and Oracle – stand in the way. 

Customer power in all economic spheres has been noted by forward thinking observers for some time – and the battle has been joined in spades in the SAP community. SAP has been facing a revolt among its users over what it charges for maintenance, one of its major revenue streams. Its delivery of a more gradual, less invasive update process paradoxically complicates that further. It’s hard to convince people to pay for the right to refuse granular updates they might not think they need. So far, only some 3500 customers have taken advantage of the new “enhancement packs” = which also are only designed for the newest release. 

Responding with new technology in the face of that recalcitrant community is a major challenge, but SAP knows it must deal with such model breakers as cloud computing, mobile devices and what it referred to throughout as “on-demand” (SaaS). Showing slides about how rapidly technologies are adopted made for an unfavorable comparison to those customers not even on current releases yet. Will the innovations on display help move customers forward? Only time will tell, but SAP is soldiering on.

SAP is very upbeat on in-memory software – claiming that access to databases in main memory is 10,000x faster than from “storage drives” – which academics like Daniel Abadi of Yale say is two orders of magnitude too high. But there’s no question it’s a big step up nonetheless. And SAP has begun to deliver “on-demand” with Business By Design (with other offerings arriving soon and more to come. Mobility is a key theme: even the Droid was in evidence at this event, scant weeks from its public release, earning a memorable “double nerd points” score in a tweet from James Governor. Oddly, the Blackberry, so prominent in the past, was absent in the sessions I attended. (More on the tech additions below)

John Schwarz’ experience at the helm of Business Objects before the acquisition showed in his savvy acknowledgement that “We need to deliver with people who understand the customer’s business,” so specialists in the sales force are critical. This is one of SAP’s powerful advantages – it understands the businesses it’s in very well, and can leverage that if it executes well. Schwarz pointed out that “We have 7000 go to market partners,” and 2.5M developers plugged in. The ecosystem work done for the past few years by the recently departed Zia Yusuf will be another key asset. But no actual results were described, just membership statistics, other than some onstage partner anecdotes.

Oddly, the acronym “GRC” for governance, risk and compliance was rarely heard. SAP has achieved good recognition there, but at this event “Sustainability” and managing risk and compliance were conflated into a new message. SAP believes there is a 7B euro market in this newly defined space in the next few years, and has a panoply of offerings poised to exploit the needs of companies confronting new legislative requirements. The already signed customer list was impressive, and Peter Graf’s demos of the technology were coherent and interesting.

There was much talk in general sessions of the “heterogeneous landscape” in the customer and prospect base. But SAP said little about connecting its apps to non-SAP apps. It plans to “orchestrate over a heterogenous landscape,” said Becher, but this seems to mean SAP is the point of orchestration across end-to-end processes it owns. Snabe referred to an “SAP fabric to which new pieces – from SAP, not others – could be added. “We believe in horizontal integration across business networks,” so we don’t need to vertically integrate hardware and databases, he added, in a jab at the competition’s different strategy. But why follow this thrust with the enthusiastic discussion of – wait for it – in-memory database? And pictures of blades as a key enabling technology for the planned innovations? Mostly, SAP’s message said much less about openness: despite Oracle’s “full stack” story, it talks a better game about moving data, messages, and events across other vendors’ products where the customers have them.

The exception to this came in EVP and GM Marge Breya’s discussion – call it the North American wing’s vision. She and Vishal Sikka, and Marge Breya discussed the “leaning down” of elements like app servers for ABAP to provide improvements that were as non-disruptive as possible to existing customers. While the embedding of BI into SAP’s own apps continues to be a strong and effective focus, Breya emphasized its continuing focus on supporting non-SAP data, albeit in read mode for analytics. SAP claimed to be twice the size of its nearest analytics competitor – though the tweetstream bubbled up some challenges to that assertion. SAS, Oracle and IBM would no doubt want to construct very specific collections of numbers to contest the terms of the leadership definition.

Breya is intent on driving continued development, and showed some terrific demos (herself, as is her style – kudos to her for that.) We saw Explorer running on iPhone – searching for a brand’s results, charting them, drilling down by geography, charting again, filtering again into specific products , and finally to the top 5 dealers. All on an iPhone, with a simple, clean interface.  Kona, the next generation BI on demand environment, will have on demand, on premise, on device delivery, and enhanced visualizations, she said, and showed a word cloud and a heatmap inside a portal during a demo of picking the best cities for a marketing campaign.  She talked about marrying middleware to analytics, as we discussed in our interview in April,  and made it clear that she has not abandoned the Orchestration roadmap – marry NetWeaver to BO,  bring up a CEP engine and incorporate ALM in the medium term. She promised collaboration and operational BI features, and suggested that delivery midyear is possible.

The fundamental assertion of the day was clear: “the foundation has been laid. Now we can accelerate delivery.” SAP’s embrace of in-memory will mean it will change its entire stack, its application servers and provide elastic combinations and multi-tenant support. These are formidable transformations for the 92,000 customers iy has after 37 years. 100 of them have been with SAP for 30 years. “And everything changes but relationship,” Snabe said. Is that enough? Execution, and the willingness of its customers to embrace the changes, will tell the tale. The future is uncertain.

Xkoto, the database virtualization pioneer, has generated substantial interest since its first deployments in 2006. Still privately held and in investment mode, Xkoto sees profitability on the horizon, but offers no target date, and appears in no hurry. Its progress has been steady: in early 2008, a B round of financing led by GrandBanks Capital allowed a step up to 50 employees as the company crossed the 50 customer mark. 2008 also saw Xkoto adding support for Microsoft SQL Server to its IBM DB2 base. Charlie Ungashick, VP of marketing for Xkoto, says that 2009 has been going well, and the third quarter was quite strong. And at the end of September 2009, Xkoto announced GRIDSCALE version 5.1, which adds new cluster management capabilities to its active-active configuration model, as well as Amazon EC2 availability.

“Traditional” models of passive failover and passive disaster recovery are high cost, inflexible architectures that xkoto’s GRIDSCALE replaces with multiple identical databases copies that it manages in an active-active configuration for lower cost scale-out and disaster recovery (DR.)  Applications don’t “see” it; GRIDSCALE captures the SQL statements and replicates them to the copies – both DDL and DML – with an optimistic, non-synchronous protocol: the first successful response goes back to the application.

Customers include CNN, Puma, HSBC, and the US Department of Homeland Security. Ungashick says Western Europe is doing very well – half of Q3 revenue came from Europe. The firm has recently added direct sales headcount in South Africa and Asia, and is continuing to add more. The partnership with IBM has been instrumental in some big wins for both parties. Xkoto is arguably the closest DB2 has to answer Oracle’s RAC, and Xkoto participated with IBM in several deals that needed this capability.

GRIDSCALE version 5.1’s Amazon EC2 availability enables multiple DB2 databases running in the cloud to work together, and avoid the limitations formerly resulting from the lack of shared storage – allowing load distribution in the cloud for the first time.Version 5.1 also adds automatic recovery, Kerberos support for authentication, and other features as described here.

The SQL Server version of GRIDSCALE 5.1 features a new, “driverless” configuration. Native SQL Server drivers, including ADO, .NET Framework, and OLE DB are now supported; GRIDSCALE has implemented the tabular data stream (TDS) protocol Microsoft inherited and updated from Sybase. Microsoft SQL Server Enterprise Manager and other tools compatible with Microsoft interfaces can be used for management of the server instances. Ungashick says he’s seeing more opportunities with Microsoft where the competition is Oracle RAC, similar to what Xkoto had already been seeing with DB2 prospects:

A number of situations have arisen recently with SQL Server customers recognizing that their data warehouses are not adequately designed for availability and disaster recovery. As the DWs become more important to the business, we think we’ll see much more interest in that use case,” he said. 

Xkoto is hoping to leverage the Microsoft community to drive business there.  The recognition the company has already received – Best of Microsoft Tech•Ed 2009 and Gartner‘s “Cool Vendor in IT Operations and Virtualization” among them – will go a long way towards boosting its visibility. This is a promising model, and Xkoto has the early lead – which will be a challenge to hold as the big database vendors add their own capabilities in this area. Meanwhile, Ungashick says, there are other database products that could use similar capabilities and we can expect to see announcements with others in the year ahead.

Many SAP and Oracle apps customers would rather leave stable products alone than continually change, or “upgrade,” as it is called. For these customers, the cost of maintenance, also known as “buying it all over again every 4 years,” seems excessive. The slow pace of innovation from the mammoth firms, and the even slower uptake of those innovations, amplifies this. (For a recent discussion of this problem, see video highlights from Ray Wang’s keynote speech from the SAP UK and Ireland User Group. I discussed the resounding thud heard from Oracle’s “wait till next year” non-announcement of its Fusion apps here.)

With this backdrop, Rimini Street, one of the pioneering 3^rd-party maintenance firms, recently announced stellar Q3 results: revenue up 200% year over year, and sequential quarter-over-quarter growth continuing: it claimed Q3 invoicing doubled the prior calendar quarter. Rimini Street’s value proposition has steadily attracted customers willing to try a different way. The company claims hundreds of customers since inception, all over the size spectrum. The offer is simple: their base price is 50% of the vendor’s maintenance price.

Rimini Street also offers premium support for sites that have customized the products (and most have), so even wary prospects may want to take a second look. Most customer issues that require attention, the company claims, are fixable. They assert that the average Rimini Streeter has 10-15 years of experience, and each account gets a dedicated engineer. Rimini Street offers a service level agreement of 30 minute response time, and claims an average of 4 minutes. Siebel, PeopleSoft, JD Edwards World and OneWorld are all offered in addition to SAP. Specifics can be found here . In its quarterly results press release, the firm claimed over 90% customer retention. Rimini Street says those who don’t renew are not dissatisfied, but rather have moved to next generation products it doesn’t (yet) support.

 In late September, Barron’s reported that Adams Street Partners had invested in Rimini Street for a minority stake, and the funds will be used to grow European operations. Rimini Street has also brought in ex-Sybase CFO and SVP of EMEA operations Pieter Van Der Vorst as its new CFO. In 2010, the firm expects to grow from 130 people to 200 – and to continue to grow its existing footprint in countries like the UK, Germany, Netherlands, Singapore, Australia, and Brazil.

Many of Rimini Street’s customers won’t be named. The vendors being displaced do not look kindly on competition, and some customers are concerned about their response. But Pepsi, Virgin Mobile, and JB Hunt have all been identified, along with extensive coverage of Siemens’ decision to use 3rd party maintenance, and Rimini Street’s website continues to show more client logos as more firms and governmental clients become willing to acknowledge – and praise – their relationships. (Perhaps the EU ought to focus on real competitive restraints that big apps vendors have attempted over the years – to customers’ material disadvantage – here, rather than on unrealized problems that might happen with acquisitions, as it’s doing with the Oracle/Sun deal.)

Risks? To be sure. There is no current litigation against Rimini Street, although Oracle has attempted to bring it into the case against TomorrowNow, which is still in the courts. Seth Ravin, Rimini Street’s founder, is widely regarded as extremely knowledgeable about the nuances of contracts, and the company is confident that legal challenges will not become a problem; it is incorporated in Las Vegas, regarded as fairly unfriendly to the kind of challenges it would be expected to see if Oracle and SAP decide to up the ante. Still, as Rimini Street continues to eat away at a key revenue stream, we may see some fireworks soon. Continued SAP and Oracle customer dissatisfaction with maintenance pricing, even as the two giants grow their margins by raising prices for the “locked in,” is likely to drive solid growth in the year ahead.

Oracle has high expectations for its newest release (Oracle Database 11g R2.) “We expect 45-50% adoption of R2 by next year,” said Mark Townsend, Vice President of Product Management, at the database analyst day during Oracle Open World recently. Such a rate would be unprecedented, but Oracle has good reasons for its optimism. Many new features target extending cost-effective use of the systems (server, storage and software) in place, and the financial drumbeat was clear and consistent. Many of these benefits are due to Oracle’s increasing ability to leverage organizations’ architectural tiers: smarter use of  interconnected servers, storage, and memory are driving performance improvements at many levels. Should Oracle’s acquisition of Sun win through, one can expect to see an acceleration of this trend.

 By now, Oracle RAC is well established (Gartner reported 15,000 companies using it as of the beginning of 2009) as a way to leverage middle tier commodity server hardware. “When RAC arrived with Oracle 9i, it was seen as a high availability (HA) play,” Townsend said.

 With 10g we saw people unlocking the value of smaller servers, but it was still typically one app per cluster. Now we see big customers stand up a grid to consolidate mission-critical and departmental apps onto one cluster.”

Oracle reports 40-400 databases in the wild on a single cluster. And whereas today many customers still buy the biggest box they can, SMP boxes have a” tax” associated with them. Logical partitioning of big boxes is a typical practice; 11gR2 introduces policy-managed dynamic cluster partitioning via Server Pools, which can be assigned to take similar workloads. Oracle can dynamically adjust the boundaries of the pools, and the promise for the future is to enable adjusting them based on performance, not only on failure. Prior to R2 it was possible to do so within a database; now it is possible to do so across them. It may not seem obvious to critics (and they are legion) of Oracle’s pricing – especially on maintenance – but these are ways to save money. Although of course in many ways they shift spending from other vendors, not Oracle itself, an even more intriguing play.

RAC OneNode offers another opportunity: an additional way to tackle consolidation with a single instance of RAC running on one node in a cluster. Oracle calls its strategy a “Maximum Availability Architecture” – when a customer adds redundancy, the database can use it for added performance. Active Data Guard lets customers move backup and reporting to redundant machines. It’s also useful in maintenance scenarios, reducing the hardware requirements: temp start a second instance, do maintenance and then roll back (this was referred to as an “Oracle Motion type of event.” Increasingly, this is an “online everything” story. Apply patches across the  cluster, do full db/OS upgrades across cluster, online reorg – the opportunities are legion. One of the most intriguing features is online application upgrades. Planned downtime is always a hard problem – organizations often need to stand up an entire parallel environment for such changes. Instead, it is now possible to make code changes in a new “Edition.” Redefinition views project to a  new version only seen by its users; triggers put updates into both places to ensure updates are replicated. “All Oracle’s applications are instrumented to use this,” we were told. An 8-node RAC cluster is the typical configuration, but some 16s are showing up, Oracle says.

Leveraging storage is the next key opportunity. “11gR2 now is a general purpose cluster file system,” Townsend asserted. Storage growth is not about to slow down, and Oracle is targeting the issue by targeting storage virtualization as a means to lower costs. Automatic Storage Management helps break down the divide between sysadmins and DBAs; it implements the “stripe and mirror everything” practices in wide use. I have not looked at the relative cost numbers yet, but Veritas replacement is clearly in Oracle’s sights here. The claim is that over 50% of Oracle’s customers on 10g and 11 adopted ASM, but it didn’t have binaries and external files. Now it will, extending its value. Customers tend to map system files to types of storage based on performance requirements – but that means year by year, they still buy both (cost) levels of storage. “In any system only about 5% of the data is active,” Townsend says, “so we’re enabling Oracle to move the less used data over time to less expensive storage. With 11g we can use advanced compression and get 4x compression – giving customers half of their disk back.” We’ll need to see what the performance hit is with real customers, but this is clearly another promising value proposition for Oracle to sell from.

At IBM’s 8^th annual Connect meeting with analysts, Steve Mills, Senior VP and Group Executive, had much to crow about. Software is the engine driving IBM’s profitability, anchoring its customer relationships, and enabling the vaulting ambition to drive the company’s Smarter Planet theme into the boardroom. Mills’ assets are formidable: 36 labs worldwide have more than 100 SW developers each, plus 49 more with over 20 – 25,000 developers in all. Mills showcased all this in a matter-of-fact, businesslike fashion with minimal hype and little competitor bashing. A research project aimed at extending Hadoop usage to a broader audience was among the highlights. 

Mills gave us a look at his organizing principle:

We have been working on extending the notion of what middleware is. It’s about connecting an organization’s applications, the codification of business process and function.”

Companies from midmarket to large enterprises run thousands of applications; understanding customers’ business scenarios, addressing identified gaps and promoting recommended patterns for success - adoption routes, solution stacks - is the driver. “It’s very easy to make a mess if you’re not guided,” Mills points out. He’s an effective, dedicated proponent of IBM’s Smarter Planet theme, and returned to it at this event, pointing out how IBM-supported projects that instrument and enhance the world’s often aging physical systems pay for themselves in efficiency savings even before the larger goals they enable are considered. He also held forth on other favorite topics: Industry Models, Cloud Computing (“You have to talk about Cloud”), and more, but told us he’d promised not to use all his team’s best slides before they could. “Not that I can’t talk about all of it,” he joked – and we’ve seen him do it. But no 3 hour keynotes here, mercifully, unlike some other vendors’ recent events.

Bringing Hadoop to Business Users

In his presentation, Rod Smith, VP, Emerging Internet Technologies, made it clear that the company is not ignoring the MapReduce/Hadoop phenomenon. He referred graciously to Cloudera’s work and picked up their phrase: “big data.” With the world creating nearly 15 PB of new data per day, a new class of content-centric WebApps is on the horizon, typically “longer running apps” – customers Smith talks with don’t like the word “batch,” he noted. But his focus was different from other vendors I’ve been hearing, where there is an assumption that the “big data” opportunity is limited to the sophisticated programmers who have so far led the way. Instead, “Put the business person in the center of the data,” Smith suggested. ”They want their own Google” – here meaning not a search engine, but a data interaction tool capable of visualization and other forms of manipulation.

It’s clear that the need for such solutions will be there, and someone will fill it. When a firm like Extrabux can process 40Gb/day, loading and indexing 70 million constantly changing input records for MapReduce by processing on Amazon’s EC2 cloud for less than $5000 per year – with no DBA – others will follow. (See the September issue of Charles Brett’s Insight-Spectra  for details of this case study.) Like other explorers in this new mode, Smith offered his own great examples, including  a Visa risk modeling app using Hadoop with the R statistical libraries that reduced an analysis literally from 1 month to 13 minutes. “This is not incrementally better; it changes everything,” he said.

Smith’s Big Sheets project showed off analysis performed on over 2 million patent documents – a “one person project, like all my things.” He referred to the iTunes interface and showed a similarly clean, intuitive model. And he pointed out that “the data operated on does not always get reduced; here it exploded, because one analysis was of how patents made references to other patents.” Similar things happen when analyzing social graphs; it’s why focusing on MapReduce alone to describe these cases doesn’t always paint the full picture. It’s just one step in more complex processes that can be distributed around large systems which scale on demand as needs dictate. Similar thinking about user empowerment, without the elastic scaling (yet), is behind Microsoft’s PowerPivot, which treats Excel as the UI, and adds operators to the Excel language which mimic the kinds of things MDX programmers can do with OLAP cubes, among other things.

IBM is looking past today’s MR cases, which are often reminiscent of early computing days, when specialists spent days to set up machines for a single program run. The problem then was scale too, and learning how to use machine resources efficiently was job one. Today, the economics have flipped – we understand that the people resources are more valuable and we have to empower them. IBM is looking beyond complex setup, java coding and single run models for “big data” processing and towards interactive big data analysis – at Web scale. In Smith’s view, that’s the key to going into an “evidence-based business world.” IBM is focused on hiding the complex details of system parallelization, fault tolerance, load balancing, etc. from the user by hiding everything behind the UI. Tech details weren’t at the top of Smith’s agenda for this presentation, but REST interfaces, the use of Jackal, extensibility via UDFs, integration of Pig, and exporting results into feeds and XML were briefly highlighted. As IBM continues to push at this area, we can expect to see some breakthrough innovations emerge, in larger, end-to-end scenarios.

Sybase has announced yet another record revenue result for the third quarter of 2009.  Like other leading data management firms, its database business demonstrated continuing vitality in a difficult economic period. With 32% growth in database licensing revenues against a strong year over year comparison, the venerable DBMS provider continued a string of recent strong results.

A 30% operating margin was icing on the cake; John Chen has proved to be a careful steward for the firm as it has carefully navigated a changing environment for its database, mobile middleware, and messaging businesses. Although Europe is still not strong, Sybase’s steady inroads in Asia, particularly China, continue to be a spur to growth, and John Chen called out the strong US Federal business as a key contributor as well.

Chen was upbeat about the future, indicating that the pipeline looks good and raising the guidance for Wall Street to suggest that the company will achieve another record in Q4. Expect to see a  little more presence from the surviving “tier 1.5″ database vendor – but not enough. Sybase continues to hide its light under a bushel, ceding too much mindshare to noisier competitors. Perhaps in 2010 we’ll see a rampup in visibility, but I’m not counting on it. And with these results, maybe they’ve got it right. Still, one can’t help but wonder: what if they made some noise?

How good was Teradata’s Q3? Not bad, but no improvement over a so far lackluster year, which nonetheless has seen the stock  price rise steadily. In 2008,  the striking rise in Teradata’s Linux revenue growth was matched only by the corresponding drop in its Unix revenue, and that “steady as she goes” performance continues through its still unevenly applied OS transition. In Q3, revenues were down a little (3%) year over year, and margin was flat (down 0.6%). YTD product revenues are down 11%.  Service revenues were up 5% for the quarter but only 2% YTD.  Still, net income rose 5%, in part because of strong expense controls. Since early 2008, Teradata has lost a little momentum through a difficult economy compared to its rivals at Oracle and IBM. Its next transition – after independence from NCR and the OS shift – is a product portfolio change catalyzed by the growth of appliance competitors like Netezza. So far, Teradata has managed to drive the product changes into the market well, claiming 65% of its appliance sales are new names. The hot new all-SSD Extreme Performance Appliance is now coming onstream, and will create a new category advantage if, as Teradata believes, there are customers willing to pay for its spectacular performance.

Q4 may be a challenge, and will be compared to a strong Q4 that helped last year’s numbers. ”Predictability has been a challenge,” management said on the conference call – but they raised the guidance. Why? Encouraging factors include the growth in sales territories as they have invested in their go to market engine. It always takes time to achieve momentum with new sales teams, so there is reason to be optimistic that an improving economy and more feet on the street will begin to make an impact. But that is by no means certain, although new account wins have been “steady across Q1, Q2 and Q3″ this year, as CEO Mike Koehler said in response to a financial analyst’s question. He said Teradata was not feeling competitive pricing pressure or sales cycle lengthening. Still, the impact so far has not been enough to prevent the revenue declines.

 Teradata doesn’t give numbers of new customers, but named quite a few marquee ones, including Comcast. Vonage. Habitat for Humanity, Royal Bank of Canada, Discover Card, and Aviva. And although competitors have been touting wins in Teradata shops, the company continues to get substantial upgrades and new business from many of the same names. Firms like Netflix, Vodafone, Electronic Arts, Qantas, Westpac, Overstock.com, eBay, ETB, Telefonica Argentina, the  US Defense Logistics Agency, Pakistan Mobile and the Commercial International Bank of Egypt reflected global success, although Asia was a weak link .

Koehler touted the recent Partners conference and its customer presentations as a showcase for Teradata’s success. My own experiences there certainly agree with this assessment. A loyal base making transformative applications with Teradata technology is one of the company’s strongest assets. Teradata announced cloud availability on Amazon EC2 for developers to create their offerings, and VMWare availability; both of these expand access, but not much revenue.

On the product side, the story is emerging, but needs a little polishing.  Teradata now has 5 platforms in its “purpose-built product family;” Koehler listed them all on the financial call. This was overkill. Financial analysts are less interested in those details than in the positioning. Similarly, in his Partners conference keynote, the unfinished status of the positioning was very clear – Koehler gave a strong, well-articulated rationale for the top of the line EDW, with reference to Teradata’s workload management, one of its greatest strengths. But the transition to “We like data marts too” was jarring and seemed flatly contradictory – the story needs to begin with the assertion that there are many use cases that require differing configurations, “and we have them.” (Note that some commentators are skeptical that workloads are homogeneous and predictable enough to warrant such configuration specificity; Curt Monash discusses the issue here.) With its strong base of happy customers, Teradata can back that story with real world examples better than many of its competitors – its customer reference organization is to be commended for its exemplary effectiveness.

The product lineup is shown here, from a slide used to brief analysts. Reading left to right will no doubt be confusing. One hopes the presentation will change as the story develops. I’ll discuss them in a different order below. From my perspective, the most exciting innovations here are in the “Extreme” products, and I’ll discuss them at the end.

The Data Mart Appliance (551) is a single node, single cabinet offering with an upper limit (it has no interconnect) of 6 Terabytes, running on Windows, SUSE Linux or Teradata’s Unix MP-RAS. The Data Warehouse Appliance (2555) handles 20 Tb and four nodes per cabinet but can scale up to 11 cabinets (half cabinets are available for more granular scaling needs) for up to 230 Tb of usable data and for some reason runs only on Linux. It uses the Teradata BYNET over Gigabit Ethernet and adds automated node failover (hardly needed for the mart since it’s only one node.) At the top of the line is the Active Enterprise Data Warehouse (5555), with 9 nodes and up to 300 Gb per cabinet and a maximum of 1024 nodes and 10 Pb. It restores Unix MP-RAS and Windows to the OS list and the interconnect is the 3.2 version of BYNET. Teradata extends its availability with hot standby and a dual active system capability.

The big bet is next: Teradata has identified a use case that it calls “hot data” for operational BI and has taken the leap ahead into solid state disk (SSD), asserting that those customers who need its extraordinary speed are ready to pay for it. The current generation of SSDs uses interconnects and form factors that mimic disk drives, and pay a performance price for it, but the improvement is so great that it doesn’t matter yet. It changes the physics: over the past decade the performance of processors has moved farther and farther beyond that of disk. Using memory-based stores completely transforms the art of the possible. They permit aggregations of millions of rows in a second, and load speeds hitherto unheard of. The hot new  Extreme Performance Appliance (2550), now in beta, supports 9 nodes and 9 Tb per cabinet, with an upper limit of 24 of each. It runs only on SUSE Linux and uses the BYNET interconnect over 10 Gb Ethernet – the only configuration to do so (so far.) On the Partners floor, this was a sensation – everyone wanted to see “The Blur”, and the buzz was everywhere.

The other Extreme is the Data Appliance (1555) - neither mart nor warehouse. It scales beyond even the Active EDW with an upper limit of 5Pb, using 9 nodes per cabinet as that model does, and the same limit of 1024 nodes. It uses BYNET V3 but doesn’t have all the availability features of the Active EDW, or the latter’s support for Unix MP-RAS and Windows – it’s SUSE Linux only. The storage difference comes from its use of 1TB SATA drives – up to 124 of them per node. This machine is pointed at the “big data” market, which often uses data that is not considered data warehouse fodder – new data types and archived data that is used for advanced multiyear analyses, for example.

The future laid out for us by the Teradata team has a solid roadmap with exciting possibilities. You can get a great summary of some of the issues and opportunities in an exeptionally useful post from Curt Monash here. But one hopes Teradata rationalizes the bewildering story, the wild mix of OSs, interconnects and disks, into a more coherently told story. Sales forces respond slowly to changes of this magnitude, and the confusion during this transition will not help. To its credit, Teradata has introduced fascinating tools for the assessment of customer needs that will help, and if it executes on this transition as well as it has on the past two, Teradata will remain formidable in the next few years as this cycle matures and garners customer successes.

Oracle is the company that led the industry into making RDBMS the data persistence vehicle of choice, and though its flagship is still Number One, many other topics floated around as 35,000 people attended Oracle Open World (OOW) in San Francisco recently. The spotlight stayed firmly planted: “What will Larry say about clouds/IBM/Fusion apps?”; Marc Benioff and Larry; Arnold and Larry. But if there’s anything Larry Ellison is passionate about, even as he sets his sights on IBM (hardware) and SAP (apps) – his two most important competitors, he said at the Churchill Club recently – it’s database, and he’s energized by the appliance opportunity. Andy Mendelsohn, SVP of Database Server Technologies put it simply in a conversation: “the only product Larry has spoken of in the last 3 earnings calls is Exadata.” He is more involved than in recent years, and that means one thing: everyone else had better watch out. What analysts learned about the new release makes that very clear: Oracle has been busy, and there is a lot of exciting new technology coming.

Still, at OOW the main tent venues, Moscone North and South, were filled with applications speeches, apps partners on the massive show floor, and apps customers. No database. Larry used the event’s first keynote on his comedy routine with Scott McNealy, bashing IBM. The next day, Chuck Philips and Safra Katz squandered their Monday keynote on awkward handoffs and disconnected stories. HP’s Ann Livermore followed with a sales pitch that sent throngs to the exits. But on Tuesday Thomas Kurian nailed the “full stack” story, despite the odd staging decision to invent a fictitious customer for demos and invite real customers on stage to comment on it, instead of discussing their own successes. Kurian’s pitch became overlong, but was well done otherwise. Then came Michael Dell, who copied HP’s themes – we’re green, we’re saving money with consolidation – and got  the same result in massive crowd departures. Larry ambled onstage for a brief appearance to thank Michael, and pointed out that Oracle has 20,000 Dell servers. I concluded that Michael can now model his lost revenue as the replacement cycle rolls on and Sun rolls in. On Wednesday, Larry returned, and spent a lot of time on his Exadata pitch, the Governator, and some vague armwaving about Fusion apps.

The usual database hyperbole was in evidence: banners like “the first OLTP database machine” are laughable, as is “first flash-optimized machine.” As always, there is plenty to take issue with in Oracle’s claims and headlines – but the quibbles will be forgotten. Larry says it loud, and it gets remembered. That’s what matters at events like this – few checks are written for purchase on the spot, but the impressions are made, mainstream headlines appear, and the engine rolls on. A block away in Moscone West, there was another set of presentations, partners, customers and competitors talking about what made Oracle the behemoth it is today. Data. Database. Database tools. And a well populated show floor to match the floor show we’d seen on the main stage.

For the database analyst community, there was also an effective and rich day of presentations off in a nearby hotel. And that was time very well spent. What was covered? There’s a new flagship release: 11gR2, which shipped  9/1 on Linux, will arrive on most other platforms in Q4 2009, and will run on Windows next year. And of course, there’s v2 of Exadata: the sequel (pun intended), featuring Sun machines as HP gets shoved off the boat. V2 starts to ship at the end of October, and versions with 2Tb SATA drives will be available in December. The positioning of all the database versions is about lowering costs, and it was consistently hammered home throughout the day to us as much as it was to conference attendees. There was a lot of tech talk too, from a substantial array of new features and enhancements, deployed in various product versions from RAC to EE to Exadata. Here’s just a taste:

• Hybrid Columnar Compression
• Smart Flash Cache
• High performance in memory parallel query (PQ)
• Online Application Upgrades
• Real time SQL monitoring
• Simplified RAC install
• RAC One Node – packaging RAC stuff but at lower cost
• RAC support for Oracle VM
• ASM cluster file system
• Secure backup (don’t use Symantec anymore)
• TimesTen (PL/SQL support)
• Berkeley DB
• Enterprise  Manager
• App Express and SQL Developer 2.0
• Audit Vault support for DB2 and Sybase
• Database Vault (for SAP), Transparent Data Encryption for Oracle Apps

This is a huge list, and blog posts that get much longer than this one are too long. Other posts will tackle some of these topics, but not all. I will comment soon on the recent TPC benchmark result Oracle is talking about – they’ve been negative about TPC for a while, but now they have one to tout, and the hyperbole reaches its apex in that discussion. If you want to be sure I tackle a favorite feature/function of yours here, please leave me a comment and I’ll do my best to get to it soon. But the net takeaway for me was this: Oracle is treating this as a very important release. They’re locked and loaded. It’s going to be a fascinating year.