News Stay informed about the latest enterprise technology news and product updates.

Sun contemplates its next Web services buy

The451 believes that Web services start-ups Systinet and AmberPoint, founded by former Sun executives, could be too attractive for Sun to resist.

Sales are down and the share price keeps sliding, but Sun claims it's starting to win deals against IBM with a new customized sales process. Most strikingly of all, executives say the company is still interested in acquiring technically savvy startups to flesh out its product line. An acquisition by Sun isn't the financial plum it might have been in 1999 or 2000, but it's still better than going under.

Context: Technology developed elsewhere has become increasingly important to Sun over the past four to five years. Consider its Sun ONE product family. The Web server portfolio formerly known as iPlanet is the legacy of the Sun-Netscape alliance, itself a side effect of the acquisition of Netscape by America Online in November 1998. StarOffice traces its origins to the August 1999 acquisition of StarDivision. The Sun ONE Studio is a collection of tools inherited from Forte Software (August 1999) and NetBeans (October 1999). Sun Grid Engine came to the company via the July 2000 Gridware acquisition. Most recently, in February 2002, Sun added Clustra for its high-availability datastore.

Java and Solaris aside, there's very little in the company's much-vaunted middleware stack that actually originated within Sun. Even the Solaris 9 Resource Manager, of which the company is inordinately proud, is a rewrite from scratch of Aurema's ShareII scheduling engine. JXTA, the pet project of Sun's chief scientist Bill Joy, is based in part on technology acquired with P2P pioneer Infrasearch in March 2001.

On the hardware side, the September 2000 acquisition of Cobalt Networks has become the basis of Sun's shiny new x86-based LX50 Linux servers, while the July 2002 purchase of stealth company Afara WebSystems seems to point to a low-end Sparc line.

Financial: Whatever its faults, Sun can't be accused of suffering from 'not invented here' syndrome. With the notable exceptions of Netscape and Forte, its typical target is a privately held vendor of critically acclaimed technology. Sun tends not to buy market share or customer traction so much as intellectual property. The company apparently has considerable confidence in its own ability to shift boxes – confidence that might have seemed misplaced this time last year, but that appears somewhat less so after 12 months of concentrated effort to focus the marketing message and simplify the product line.

It's much too soon to say that all will be well with Sun. High-profile executives have left the company, sales are expected to come in about 15% lower even than last year and many observers are concerned that in spite of these hard times, operating expenses will rise. IBM is out to eat Sun's lunch, and it's been enjoying a certain amount of success.

All these factors are reflected in the company's pathetically deflated stock price: shares are around $3.50, from a 52-week high of $15 or so, far below the glory days of $60-70 when Sun put the dot in dot-com. Even so, the terms of Sun's refocus – an emphasis on technical excellence, and especially a customized sales process said to be closing deals even against Big Blue – suggest that its acquisition strategy will remain intact. Look for the company to keep shopping among early-stage startups with unimpeachable technical chops.

Strategy: Just where might it find them? One indicator of Sun's future interests is, ironically enough, the subsequent careers of executives who have left the company. You might say that the giant vendors, and especially Sun, have outsourced part of the risk of research and development to veterans who helm startups in promising sectors. If the startups succeed, Sun or one of its competitors can reap the gains without running the risk. Of the current crop of ex-Sun vehicles, three in particular stand out.

Ejasent, founded as Apera, is the brainchild of Rajeev Bharadhwaj, the former architect of Sun's JavaOS. He remains on board as CTO. The company's software dynamically assigns servers to multiple applications, letting data center operators improve utilization and billing. It's a key player in an emerging resource-pooling market that also involves Jareva, Terraspring, Think Dynamics and even HP and IBM.

Systinet, which started life as Idoox, is the fifth venture from Czech Republic entrepreneur Roman Stanek. His fourth startup was NetBeans. Stanek was still working for Sun when he rounded up a fresh team of talented developers and set them to work in Prague, this time on XML and Web services. Their pioneering SOAP engine won rave reviews and earned the company over $23m in venture financing. A competitor to Cape Clear and The Mind Electric, Systinet now boasts a remarkable 5-million-user deal with Europe's T-Motion wireless network, and wields an influence in the Web services world that is well out of proportion to its small size.

Finally, there's AmberPoint, formerly Edgility, founded by Sun and Forte veterans John Hubinger and Paul Butterworth. AmberPoint's Management Foundation software tackles the thorny problem of monitoring distributed Web services – not to mention access control, online upgrades, business and system alerts and comprehensible logging. AmberPoint takes its place beside Corporate Oxygen, Digital Evolution, Infravio, Talking Blocks and WestGlobal in the brand-new but fiercely competitive Web services management market.

None of which is to say that Sun will buy Ejasent, Systinet or AmberPoint – only that former Sun superstars, including three who have already succeeded in selling their own ventures to the company, have identified resource pooling, Web services and management as hot new sectors of enterprise software. Sun has a certain amount of expertise in each area already, but it could probably use more.

The451 assessment: With HP and IBM on its back in the enterprise server market and Microsoft's .NET posing a direct threat to its own middleware stack based on J2EE, Sun needs to add competitive advantages as quickly and cost-effectively as it can. Sun -- a compulsive shopper for technically proficient startups -- will make more purchases, most likely of privately held firms in fast-moving sectors like management and Web services. Based on its past performance, the company will also do a reasonable job of integrating the intellectual property it buys.

the451 is an analyst firm that provides timely, detailed and independent analysis of news in technology, communications and media. To evaluate the service, click here.

Dig Deeper on Topics Archive

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.