In recent years, as some systems began to show their age, Choice Hotels International, Inc., which franchises over 6,000 hotels, opted to update with middleware systems from Oracle Corp. Choice Hotels’ move is well along. SOA is part of the journey.

“Several years ago, we did an overall assessment and decided there were too many point-to-point connections,” said Rain Fletcher, vice president of application development and architecture, Choice Hotels. “Maintenance was difficult, and our business needed more velocity in delivering new functionality.”

That is the background to Choice’s selection of Oracle Fusion Middleware, SOA Suite and BPM Suite, he said. Oracle application server software also provides the base for those higher level stack elements. Fletcher described Choice’s as “a WebLogic shop,” referring to the former BEA, now Oracle, app server suite.

The legacy systems were becoming “a liability in our ability to execute,” said Fletcher. A re-thinking of IT architecture was needed, he said.

The need for quick technical flexibility is intensely apparent on the Choice Hotels’ website today. It is dotted with special book-early rate offers, gift card offerings, Privilege Point member specials, downloadable iPad and smart phone apps and more. It enables bookings on Comfort Inn, Quality Inn, Econo Lodge, Sleep Inn and other familiar hotel marquees. All this functionality must be fleet and flexible, and supported by backend systems and middleware.

A big part of Fletcher’s drive is to simplify and standardize where possible. “We wanted to default to one standard,” he said. But multiple systems are a fact of life that require developers to be supple. The that is needed is what Fletcher calls “fungibility.”

“We have thirteen different systems. And I don’t want there [to be a need for] ‘tribal knowledge’ of any one of them,” he said.

How does that pan out in operations? Some complexity is unavoidable – but simplicity must be the goal. “Every application server type we have has a different patching policy and security profile,” he said. “I may have four – I don’t want any more,” he said.

SOA and the Oracle SOA Suite have been first steps in gaining the flexibility Fletcher’s organization is looking to achieve, with BPM and modeling deployments to come, he said. Developers go through intensive training in SOA. He said the effort is built around the concept of a “SOA Services Factory.”

“We started with service domain module creation, working with partners to create several high-level service domains,” said Fletcher. Looking forward, he expects to be mapping SOA services with well-defined business processes. The domains provide a framework for the services. As the services grow up they map naturally to how the business thinks, and what the business does, he said.

@SearchSOA

Increases were seen across the playing field, with perpetual middleware giant IBM gaining the largest growth margin, adding 2% in market share, and going from 30.6% in 2009 to 32.6% in 2011. Despite trailing IBM by nearly half, distant runner up Oracle also added to its market share by 0.8%, going  from 16.2% to 17% in the same period. Other companies not amongst the top five vendors lost nearly 4% of their market share.

A series of acquisitions, totaling $1.2 billion has led to an even greater concentration of the market amongst the top five vendors. Through these acquisitions suite vendors swallowed up independent companies. The result is IBM, Oracle, Microsoft, Software AG and TIBCO now control 61% of the AIM market. This figure is a 4% increase from last year’s 57%. North America and Western Europe remain the two largest regional, situated just ahead of Japan and Asia/Pacific.

Growing technology sectors include spaces like SOA governance as well as open source technologies.

A long string of enterprise application company purchases saddled Oracle with a disjointed product line that only good middleware could remedy. While it was building out some mostly Java based middleware for in-house  Fusion Applications’ suite needs, it offered the resultant Fusion Middleware tools to the wider public. 

For some reason, these Oracle Fusion stories haven’t quite come across too vividly. The reason is the controversial Larry Ellison, who is suddenly selling hardware with a very hard sell technique.

Last month at OracleOpenWorld, both stories – Fusion Middleware and Fusion Applications – got short shrift as colorful company head Larry Ellison disparaged SalesForce.com head Mark Benioff , belittled the speed of RedHat’s Linux distro, criticized the IBM mainframe and just generally gushed as he rolled out  his new Exalogic cloud in a box.

According to Ellison:

*Benioff was “offended because [Oracle's Exalogic] box was taller than he was”;

*Oracle created a new version of the Linux kernel to compete with its present RedHat-based Linux kernel because “we needed an OS… that was a lot faster”; and (clearly aiming a barb at IBM’s mainframe);

*Exalogic is “big iron without the big price.”

So let’s break down some details of the discussion.

* Let’s not argue that Big Blue’s big iron does not cost big bucks. But let’s say that the $1-million-plus tentative price tag for Exalogic does somewhat resemble a mainframe’s tag;

*  Let’s ask if Ellison’s implied definition of cloud as ‘taller than Mark Benioff’ negates the fact that Benioff’s view of cloud has held public currency for a good while now;

* But, let’s grant that Oracle’s Linux kernel may be faster on the hardware it gained with its acquisition of Sun Microsystems.

Truth is, this final point plays to a theme submerging much else these days at Oracle.

The theme : “Hardware and software, engineered to work together.”  Ellison repeated the phrase several time in Oracle Open World keynotes. It really was a pretty hard sell. So hard that there is reason to suspect that the pressure is on to really make the Sun acquisition work very quickly.

Although the Sun acquisition was a long time in production, this has to come as something of a surprise to the vast majority of long-time Oracle customers who come to events such as Oracle Open World to see what kind of software they can place on any number of hardware platforms. Oracle’s early data base success was very much due to its ability to run well on many platforms.

A Larry Ellison that was somewhat muted during his sportsman’s quest for the America’s Cup yachting trophy has emerged from recent Oracle Open World with continued invective. Perhaps not too curiously much of this rancor is aimed at competing hardware powerhouse Hewlett-Packard. Ellison has to make quick inroads in hardware – software, including middleware, may be taking a back seat for a while.

Java was one of the major sea changes in the history of application development. It brought some ease of use to objects as well as a usable set of standards for distributed computing. With word last week that the European Union had okayed Oracle’s purchase of Sun Microsystems, the possibility of another sea change must be considered.

Oracle did not enter the Java era as a middleware leader. At the time, the company was still built around its flagship data base offering–”middleware” was anything that glued your application more tightly to the Oracle DB, thank you. After missteps, Oracle came to embrace Java and J2EE. Its purchase of BEA (for $8.5 billion in January 2008) gave it much more sway in the Java space. Middleware has become a bigger part of Oracle’s planning as the years have gone by, and the Sun purchase may accelerate that.

Java was never just Sun’s–IBM, Oracle, BEA and others had influence on Java direction. But there is no question that Sun had the greatest influence on Java direction. There are more fissures in the Java community now than there were in the past–stripped-down componentized frameworks like Spring are on the rise. It will be interesting to see how Oracle responds, and how fervently it develops and markets Sun’s open-source componentized Glassfish V.3 platform, rolled out just as the lights were going off on Sun’s independent future late in 2009.

This week, Oracle is expected to disclose a more detailed roadmap for its Sun acquisition; look for coverage on SearchSOA.com, and – as always – let us know what you want to know.

“If anything, we’re going to invest more in MySQL,” Ellison said at the conference keynote. “Not less.”

Furthering these sentiments, Oracle has released a number of statements and, recently, a FAQ stating the company’s position across many of its major divisions and aquisitions.

In the FAQ, Oracle said Sun’s open-source Java application server, Glassfish, is far from in any danger.

Oracle plans to continue evolving GlassFish Enterprise Server, delivering it as the open source reference implementation (RI) of the Java Enterprise Edition (Java EE) specifications, and actively supporting the large GlassFish community. Additionally, Oracle plans to invest in aligning common infrastructure components and innovations from Oracle WebLogic Server and GlassFish Enterprise Server to benefit both Oracle WebLogic Server and GlassFish Enterprise Server customers.

 

 

 

 

As for NetBeans, the company reaffirmed both JDeveloper’s position as Fusion Middleware’s main development tool and the open source tool’s availability should customers prefer it.

Also in the FAQ, Oracle stated its position on MySQL in the same Vein as its chief.

“Oracle plans to spend more money developing MySQL than Sun does now,” the FAQ stated.

The company also said it plans to add open-source MySQL to its existing suite of database products.

Sometimes we have to remind ourselves about the obvious things. Business Process Management is about processes. SOA is about architecture. The two have been involved in a tango in recent months, as software architects work with their business-side brethren to make change happen in the organization. On one level, the dance of architecture and process is very familiar. Yet it plays out today in unique ways as you will find in our BPM tutorial.

 The product activity in July was surprisingly active. July was once a ”quiet time.” Not so these days. Some July highlights as culled from SearchSOA.com product news:

• Software AG has released webMethods 8.0 and integrated a Business Service Repository to increase business-IT awareness.
• Kapow technologies released a new version of its flagship server. In this new version, the “Kapow Mashup Server” becomes the “Kapow Web Data Server.
• KANA developed a new platform in conjunction with IBM, using that company’s SOA portfolio to address the specific needs of the customer services industry segment.

Product news sometimes helps to identify technology trends. There are exceptions, but, by and large, vendors add elements to products because customers need the new features. Recent product news on SearchSOA.com includes a look at WebLayers’ governance tool update, which matches centralizing policy management alongside distributed administration and validation, and Layer 7’s latest appliance which now monitors the performance of both services and cloud service providers and provides access controls between cloud-based services and enterprise-based assets.

 While we are at it we would point to last month’s big product news out of Oracle. In ”Oracle Fusion Middleware 11g supports SCA, JavaServer Faces development” we see both the culmination of Oracle’s purchase of BEA Systems, and an impressive display of the breadth of middleware complexity today.

 In closing, if you have grown accustom to Representational State Transfer [REST] architecture, and are wondering what will be next on the transactional front you may want to check out Eric Newcomer’s Ask The Expert discussion on ACID and BASE. BASE stands for Basically Available, Soft state, Eventually consistent. ACID stands for Atomicity, Consistency, Isolation, and Durability. To find out more, read “Is BASE a more scalable alternative to ACID transactions?“

The suites are made up of Oracle’s existing product line and the technologies from its acquisition of BEA earlier this year, according to a brief report on the state of the current Oracle middleware offering, Oracle OpenWorld’s Middleware Message Is ‘Watch This Space,’ published earlier this month.

The Gartner analysts note that little was said about middleware in the announcements at Oracle Open World last month other than the announced plan to put Fusion in the Amazon cloud. The roadmap announced this past July for the full integration of the BEA products into Oracle’s middleware will not come until sometime in 2009, Gartner predicts.

Rather than judging the future of Oracle middleware by this interim marketing strategy, Gartner analysts recommend waiting for Oracle Fusion Middleware (OFM) 11g, due in the next six to 12 months.

That release ”will begin to implement the announced road map, and platform modernizations, such as support of OSGi Alliance technology and Service Component Architecture, expanded hot-pluggability, and the extensive use of Oracle Coherence XTP-distributed cache,” the report states.

Former BEA customers, who may be unhappy with Oracle Corp.’s plans for WebLogic as its main service-oriented architecture (SOA) server, are targeted by Sun Microsystems Inc., which today announced a migration program for its JavaCAPS SOA platform. Sun is touting the lower price and open source status of JavaCAPS for former BEA customers looking for an alternative to the Oracle version of WebLogic, said Ashesh Badani, SOA director at Sun.

The special offer announced on a new Sun Web page is aimed at BEA customers who are less than thrilled with becoming part of Oracle. It includes free adapters for integrating JavaCAPS into existing WebLogic SOA environments, guaranteed 12-month pricing, and a special five-day workshop with Sun’s SOA architects for the first 20 former BEA customers that sign up for the program, he said.

The open source status of JavaCAPS is touted by Badani as an advantage over WebLogic. Customers can avoid vendor lock-in by migrating to JavaCAPS, he argued. He stressed that JavaCAPS customers are not locked in with Sun if they become dissatisfied.

“If you’re not satisfied with us, we don’t want to make it difficult for you to leave us,” Badani told SearchSOA, adding that the onus is on Sun to keep JavaCAPS customers satisfied. “So it is up to us to really focus on customer satisfaction to be sure your deployment is going the way it should. If you think we’re failing you, you have the ability [to migrate away from Sun] – nothing’s easy – but it’s easier if your working with an open source code base rather than a proprietary product.”

The Sun’s offer is most likely to attract mid-market WebLogic customers seeking an alternative to Oracle, said Bradley F. Shimmin, principal analyst, application infrastructure, Current Analysis LLC.

“What Sun is going to do is find some new customers that have existing BEA investments and are looking to roll out a new deployment in a different department,” Shimmin predicted. “Sun will also be able to further capitalize on their existing customer base that already have Sun for operating system, storage, security, etc.”

Bringing JavaCAPS into an SOA environment based on WebLogic should not be a problem for former BEA customers wanting to give Sun a try, the analyst said.

“All these platforms are interoperable so there is no problem with a heterogeneous environment with different servers,” Shimmin explained.

Shimmin said this deal is unlikely to cut into Oracle’s overall market share with its newly acquired WebLogic product, but does allow Sun to introduce JavaCAPS, which has traditionally been a lower cost alternative, to some customers.

As Oracle prepares to announce its plans for the BEA acquisition tomorrow it may be facing a customer revolt if it’s not careful. What follows is a summary of the survey findings. The raw numbers are available here.

Who responded?

In all, we received 431 responses. Most of the respondents were BEA WebLogic Application Server and Oracle Database customers (94.90% in each case). Respondents also used a a healthy number of other BEA products (WebLogic, AquaLogic, Tuxedo, etc.), while only a small percentage used non-database Oracle products. Part of this is to be expected as the survey was geared toward the BEA user base. This group indicates that Oracle indeed bought itself a customer segment into which it had little penetration. Of particular note is that not many of the BEA customers were using Oracle’s packaged applications (e.g. financial, CRM, human resources).

Most (58.70%) came from IT shops with 250 or more employees.

Satisfaction levels

75.56 of respondents reported they were either somewhat or very satisfied with their BEA products. That confirms something this industry watcher has heard anecdotally over the years, namely that BEA customers, if not teeming in numbers, were a generally contented lot. 61.26% reported they were somewhat or very satisfied with their Oracle products. The main difference is that 24.83% reported they were neutral in regard to their Oracle products, which is perhaps understandable given that most were database customers and almost 75% of the respondents were either architects or developers, not the sort that falls overly in or out of love with a relational database.

Uneasy masses

87.94% reported they have not yet been contacted by Oracle concerning their BEA products and six months of relative radio silence has seemingly made them nervous. 29.47% reported that they lack confidence that Oracle will continue to support their BEA products and another 44.55% aren’t sure whether that will happen. Oracle has managed to allay similar fears when it has acquired packaged app vendors, but “support” in the development community will mean not only continuing service and support for existing products, but also making sure they keep pace with new advances in the marketplace. This brings us to where Oracle stands to alienate this new customer base if it doesn’t announce and follow through on aggressive plans to move the BEA product set forward (principally WebLogic Application Server).

Potential customer revolt

62.18% of respondents report they will not look to move to comparable Oracle products if their BEA products are discontinued. Another 25.06% report they are unsure on that matter. 77.26% say they do not feel Oracle has a strong offering in the areas where they are using BEA products. Additionally, 70.77% report they will look to replace their BEA installments rather than keep them as legacy if those products are discontinued. It creates a thorny situation for Oracle. It does not have a strong reputation with these, largely, app dev users and they have expressed a clear willingness to jump ship should they not like the course Oracle charts for them. While Oracle surely will look to allay these misgivings in the BEA user base, competitors just as surely will be looking to woo this potential pack of free agents.

Negative impression

Perhaps it can be chalked up to people not liking change or to unhappy customers being more likely to respond to a poll, but 52.43% of those polled reported they have a somewhat or very negative view of Oracle’s BEA acquisition. Another 32.48% voted neutral. The poll indicates that Oracle has a ton of work to do if it wants to win over these BEA customers. This is indeed a new market that Oracle could penetrate in its quest for global software domination, but these users are not rolling out a welcome mat. It may takes years of continuing and advancing key BEA product lines before Oracle can establish itself with these customers, making tomorrow’s announcement only the first step on a political tightrope that stretches beyond the horizon.

1. Which BEA products do you use? (check all that apply)

Weblogic application server–94.90%
Weblogic portal–27.84%
Other Weblogic portal—12.06%
Aqualogic service design and development tools–18.10%
AquaLogic governance tools–7.89%
Other AquaLogic products–11.14%
Tuxedo–8.12%
Other–9.98%

2. Rate your level of satisfaction with these BEA products:

Very dissatisfied—6.50%
Somewhat dissatisfied–8.35%
Neutral–8.58%
Somewhat satisfied–40.60%
Very satisfied–35.96%

3. Which Oracle products do you use? (check all that apply)

Database–94.90%
ERP/Financial-5.10%
CRM–3.71%
Human resources software–2.09%
Fusion middleware–7.19%
None–2.32%
Other–13.23%

4. Rate your level of satisfaction with these Oracle products:

Very dissatisfied—4.41%
Somewhat dissatisfied–9.51%
Neutral–24.83%
Somewhat satisfied–38.52%
Very satisfied–22.74%

5. Has Oracle contacted you about ongoing support for the BEA product(s) you use?

Yes–12.06%
No–87.94%

6. Do you feel confident the BEA product(s) you use will be given the proper support by Oracle?

Yes–25.99%
No–29.47%
Not sure–44.55%

7. If the BEA product(s) aren’t part of Oracle’s roadmap, will you look to move to comparable Oracle product(s)?

Yes–12.76%
No–62.18%
Not sure–25.06%

8. If Oracle discontinues your BEA product(s), will you maintain it as a legacy system or look to replace it?

Maintain as legacy—29.23%
Look to replace–70.77%

9. Do you feel Oracle has a strong product offering in the areas where you have BEA product?

Yes–9.98%
No–77.26%
Not sure–12.76%

10. As a customer, what is your overall impression so far of Oracle’s acquisition of BEA?

Very negative–12.06%
Somewhat negative–40.37%
Neutral–32.48%
Somewhat positive–10.67%
Very positive–4.41%