eWeek Labs Picks Best Technology Bets
eWeek Labs Picks Best Technology Bets
Technology waits for no man-or stalled economy. On the next few pages, eWeek Labs analysts take a look at the products and technologies that will be most compelling in the next year. Henry Baltazar notes that "convergence" and "interoperability" will be the watchwords for storage and business continuity as storage management software finally comes to the fore and emerging specs allow industrywide cooperation. Web services were all the buzz this year but will become a reality next, notes Jim Rapoza. And some of the most significant Web services applications may show up in the mobile space, "maximizing the potential of mobile computing," says Jason Brooks. Three little letters-X-M-L-spell big changes for the database market, according to Timothy Dyck, who hopes that a much-needed XML query language standard will be OKd in 2003. Other predictions-as noted by the Labs John Taschek, Peter Coffee, Cameron Sturdevant and Francis Chu-include, respectively, big differences in customer relationship management software, Advanced Micro Devices Inc. Hammer success, lots of security product failures and more blades.
Storage management software will remain in focus throughout the coming year, which is good news for beleaguered hardware vendors.
As in the recent past, the words "convergence" and "interoperability" will be the focus of many new technologies and initiatives that originate in the coming year.
With ratification of iSCSI nearing, IP and Fibre Channel will soon officially converge-at least in the eyes of standards bodies such as the Internet Engineering Task Force.
Storage devices running on pre-ratified iSCSI code have been around for more than a year now-with the most visible example being the Cisco Systems Inc. SN5428 storage router. The Cisco router demonstrates whats good (low cost) and bad (relatively slow performance) about iSCSI products.
The SN5428 isnt the only iSCSI product on the market. Nishan Systems Inc.s IPS 3000 Multiprotocol Storage Switch is another option, and we expect more products to become available next year. The speed gap, however, will persist.
IP offload adapters, such as Alacritech Inc.s Alacritech 1000x1 Single Port Server and Storage Accelerator, help boost the performance of iSCSI implementations but also boost the cost of implementing iSCSI. As more products enter the market next year and the powerful IP offload technology becomes more commonplace, we predict the prices of these adapters will go down.
On the interoperability front, storage management software will remain in focus throughout the coming year, which is good news for the beleaguered hardware vendors that have been hammered in recent years for the subpar interoperability of Fibre Channel hardware.
Interoperability initiatives including EMC Corp.s WideSky and Hitachi Data Systems Corp.s TrueNorth aim to make it possible to manage heterogeneous storage systems from a single management platform, but it will take a few years for these efforts to become successful.
On the standards front, emerging specifications such as Common Information Model are ramping up to enable industrywide interoperability of storage management. We expect this to be a major issue in years to come, but the emergence of these technology indicatives will be a relief to IT managers who have been demanding management interoperability for several years now.
It looks as though desktop Linux is set to make waves next year.
The stars may be lining up to make 2003 a year of significant growth and excitement in mobile computing. Whether such a constellation comes into view, however, will depend on the continued spread of GPRS (General Packet Radio Service) and 1xRTT wireless data networks-along with price schemes palatable enough for companies to adopt them.
Virtually every major wireless carrier is now in the midst of a high-speed data network rollout, and AT&T Corp. and Sprint Corp. are experimenting with the sort of unlimited access rate plans that will enable mobile device users to enjoy and afford the same always-on network connectivity thats become de rigueur for tethered systems.
As wireless data networks continue to spread in the coming year, companies can expect to see a widening range of mobile devices with which to pair them. Handspring Inc.s Treo smart-phone devices now come in flavors for GPRS and 1xRTT, and we expect Microsoft Corp.s long-awaited smart-phone platform to find its way into users hands next year.
In addition, the coming year should bring new operating system advances in mobile computing, with Palm Inc.s ARM-based Palm OS 5 due to debut, along with a decent chance that Microsofts Windows CE .Net will find its way into Pocket PC.
Advances in mobile hardware, software and connectivity alone do not a business case make. They do, however, set the table for some interesting mobile Web services scenarios. BEA Systems Inc., IBM, Microsoft and Sun Microsystems Inc., among others, are preparing mobile versions of their Web services offerings.
Mobile Web services promise to maximize the potential of mobile computing by ferrying data to and from field devices, enabling users to function more flexibly and efficiently.
I also expect desktop Linux to make enterprise inroads next year. Desktop implementations of the Linux operating system have been steadily accruing small design and functionality gains for the past few years. Helping to hamper acceptance, however, was the insistence of Red Hat Inc.-perhaps the biggest name in Linux-that the open-source operating system wasnt yet ready for the desktop.
Now, with Mozilla and OpenOffice.org settling in after their recent 1.0 releases, and the K Desktop Environment and GNU Network Object Model Environment desktop environments growing ever more polished, it looks as though desktop Linux is set to make waves next year. Red Hat now seems to agree-its just released its first desktop-oriented Linux distribution.
The next generation of blade servers will have the muscle to tackle enterprise applications such as databases.
Blade servers staked a claim in the front-end Web server market this year, with product offerings from vendors such as RLX Technologies Inc. and Hewlett-Packard Co.
However, blade sales have been sluggish as computational density, power consumption and data center real estate became less pressing in the slow economy and IT managers turned to industry-standard rack-mount servers made cheap by vendor competition.
Next year, we should see a renewed market for blade servers as server vendors including Dell Computer Corp., HP, IBM and Sun scramble to release more powerful multiprocessor server blades.
Unlike current offerings, which target only the front-end Web space, the next generation of blade servers will have the muscle to tackle enterprise applications such as databases. These high-end server blades will use the latest processors and chip sets found in current high-end servers.
As new products based on standards such as InfiniBand and 10 Gigabit Ethernet become available, server blades will be able to leverage these technologies to provide enterprises with powerful blade systems capable of handling any enterprise application.
Mellanox Technologies Inc. and other companies have released InfiniBand blade reference designs offering powerful blade architecture that can run multitier applications on a single chassis with several high-end server blades. The Nitro II server blade reference design can run a firewall, Web servers, a database and other applications on a single 5.25-inch chassis hosting 14 blades.
This year, we saw management and provisioning initiatives ranging from HPs Adaptive Infrastructure to Dells Smart IT. As we move into next year, we should see better integration and synergy between server hardware and management software.
Jareva Technologies Inc. and others are releasing complete hardware management solutions that provide end-to-end management, provisioning and inventory for the enterprise servers and blade systems.
We will see increased support for Linux in the server market next year as vendors provide more Linux offerings in their server lines. IBM has announced that its high-end Unix-based pSeries servers will run Linux natively without requiring them to run AIX on another partition. Sun has introduced the entry-level LX 50 server running Sun Linux. Although Sun hasnt announced plans to put Linux on its high-end servers, we expect to see more Sun Linux server offerings next year.
With all these obvious arguments against it, Itanium is poised to make its critics look like idiots two years from now.
Technology creators often have to make a difficult choice between protecting current users from unreasonably disruptive change or protecting future users from the errors of the past. The anger of customers today is clearly more of a hazard than the fear of derisive comments by prospective customers tomorrow. Its no surprise, therefore, that most of the errors occur in the direction of taking too long to decide that a change is due.
The coming year will be filled with more chances to make this same mistake, and even the best-considered advice may someday be seen as an example of this error. Ill take the chance.
For example, Intel Corp.s Itanium initiative continues to look like a truly bad idea. It relegates the huge installed base of 32-bit x86 code to run in an emulation ghetto, off in the corner of a chip that will devote the rest of its resources to a completely new instruction set.
Further, it hopes that the IT buyers who will need 64-bit power tomorrow havent already needed it enough to have made a commitment to existing 64-bit hardware from IBM, Sun or other providers. It demands the support of a completely new generation of compilers, debuggers, device drivers and other system software. It also seems like a terribly mistimed arrival-these days, enterprise IT is implacably hostile to anything that looks like change for its own sake.
With all these obvious arguments against it, Itanium is poised to make its critics look like idiots two years from now, when it may be that next year will be generally acknowledged as the year when Itanium turned the price/ performance corner to leave Advanced Micro Devices Hammer in the dust. With its 64-bit superset of the x86 instructions, and its road map toward continued performance improvement for 32- and 64-bit code on a common hardware base, the question is whether Hammers intrinsic hardware complexity has the headroom it will need-or if Itaniums gamble on shifting complexity to system software will pay off. Im betting on AMD.
Ill take a similar risk in my other beat, application development technologies. The coming year will mark developers first full year of access to a shipping Visual Studio .Net, and thus to Microsofts Common Language Infrastructure, with its claimed advantages over Java.
At the end of next year, its my expectation that Java will not only be still standing but also still growing in enterprise acceptance. To paraphrase Sun Fellow and Vice President James Gosling, in a conversation that we had earlier this year, sometimes language design is more about making it impossible to do stupid things than about making it possible to do extremely clever (but also quite dangerous) things.
I see more problems arising from software thats stupid than from software thats not as clever as it could be-and thats why Java still looks good to me in 2003.
Whats still badly needed is a query language natively designed for XML data structures.
Three little letters spell out the single biggest change that 2003 will deliver in terms of data management, and those letters are X-M-L.
The big relational databases already offer a number of ways to store, transform and retrieve data in XML format, with Oracle Corp.s Oracle9i Release 2 currently offering the best combination of XML preservation and relational query power.
Whats still badly needed is a query language natively designed for XML data structures, instead of having to mash XML into SQLs row-table orientation. This language is XQuery, and all the major vendors are waiting for the World Wide Web Consortium to release the standard as a recommendation, something that should happen next year.
Around that same time, I expect to see rapid adoption of XQuery across the database industry. (A few XML databases already implement draft versions of XQuery.)
These XML-driven changes will have ripple effects, such as native support for XML Schema data types (much more precise than SQLs own native types) and more powerful full-text query syntax, that will benefit all data-centric applications, not just those that work with XML. Hybrid SQL/XQuery applications will allow for a gentle technology adoption curve.
Another trend I expect to see is a shift toward closer-to-real-time and more scalable data analytics, especially on the OLAP (online analytical processing) and data mining fronts. The volumes of data being collected and the push to speed business processes are requiring faster data analysis turnaround, and that implies a shift toward relational OLAP (which can be performed directly on source data in a relational database) and away from multidimensional OLAP (where data is copied and analyzed offline).
Clustering, database high availability and stronger security will continue to be areas of vendor focus as well.
On the product front, the biggest database release I expect to see next year is the Yukon release of Microsoft SQL Server, although the post-8.1 release of IBMs DB2 might squeak in by the end of next year. Yukon will ship its first beta in the first half of next year and will be an update comparable to the huge change that happened between SQL Server 6.5 and 7.0. Yukon will finally modernize SQL Servers programming infrastructure by adding support for .Net Framework and will have engine-level XML and Windows file system integration.
The effect of Web services will be felt most deeply in large enterprise applications.
If Web services is one of the biggest technology stories of this year, it will only be bigger next year.
The effect of Web services will be felt most deeply in large enterprise applications, from portals to content management to CRM to business-to-business to knowledge management. During the last year, most of these product categories have moved aggressively to incorporate Web services, both in integrating them and in providing capabilities as Web services.
This trend will also result in increasing integration among these product areas and in outright acquisitions, as companies strive to offer full Web services platforms. This was recently seen in the acquisition of major portal player Epicentric Inc. by content management leader Vignette Corp.
This will continue next year, with some of these products essentially becoming specialized Web service management platforms. This will occur even though major divisions in vendor factions will continue to slow the adoption of important Web services standards, especially in the areas of security and unique ID management.
As for the browser wars, the introduction of the outstanding Mozilla browser will not signal a major sea change. Microsofts Internet Explorer will continue to be dominant, although many site developers will begin to code more to standards than to what works only on IE.
However, Mozilla will have a major effect as a platform for application development. The fact that the completely open Mozilla runs identically on all platforms will make it an increasingly popular choice for cross-platform application development.
As for security, look for more proactive security tools that make attacks less likely to occur, rather than simply responding to attacks that have already happened.
I also expect several major worms and cracker attacks to occur. These will be followed by increased security awareness, which will wane as company executives decide to cut back on security funding. Overworked administrators will fail to patch known problems, and new worms and attacks that leverage these problems will strike.
In other words, the same thing that happens every year will happen again next year.
In the coming year, network managers will have far more choices about when and how to throttle network traffic.
Flirting with 10 Gigabit Ethernet technology and pushing traffic management capabilities from the core to the wiring closet will be the significant networking trends next year.
These will be closely followed by a burgeoning number of security products, most of which will work but many of which will fall by the wayside as they prove too cumbersome to manage and occasionally fail in what may prove to be some spectacular flameouts.
The 10 Gigabit Ethernet 802.3ae specification was finished in June, and the 10 Gigabit Ethernet Alliance has conducted interoperability bake-offs throughout this year. Although the initial capital costs for 10 Gigabit Ethernet equipment are expected to be higher than Gigabit in the LAN, applications such as storage area networks using iSCSI could justify the higher costs.
Even so, interoperability among vendors should not be assumed, and we caution early adopters to be especially diligent when going over equipment specifications.
In the WAN world, the 802.17 working group is pushing forward efforts to make 10 Gigabit Ethernet interoperable with SONET (Synchronous Optical Network), the long-haul technology used to connect high-speed circuits. Because 10 Gigabit Ethernet provides just a hair more space than SONETs 9.9G-bps capacity OC-192, 10 Gigabit Ethernet will likely lower recurring long-haul costs, possibly late next year.
Throughout the past year, eWeek Labs has seen network equipment makers pushing quality-of-service and management capabilities that used to be found only in high-end switches into wiring closet equipment.
Ciscos Catalyst 3550 switch family is just one example.
Cisco has basically pushed Layer 3 and 4 traffic management capabilities into the switch line, thus dramatically extending network administrators abilities to regulate the who, when and how much of bandwidth allocation closer to the desktop. I predict this trend will only gain momentum next year.
IT managers will have greater control over access to the network, but with that control comes responsibility. Network and application staff will be drawn together because of the greater impact that network infrastructure devices will have on application performance.
Nearly gone are the days when simply increasing network capacity would be the best answer to solving slow application response. In the coming year, network managers will have far more choices about when and how to throttle network traffic.
The introduction of 10 Gigabit Ethernet will likely accelerate the convergence of data and voice applications, further underscoring the need for widespread traffic management.
One of the biggest ramifications I see as a result is an increased need for network planning for management and capacity. Fortunately, most of the companies that provide these products have survived the downturn while also resisting the urge to join the horde of "security" companies that emerged this year.
In fact, look for the algae bloom of security companies to die back in the middle to end of next year.
Of course, prudent IT managers must protect network infrastructure and applications, but now is the time to be choosy. Weve seen many new entrants into the security market-Reactivity Inc., for example-and even established companies change their spots, as in CacheFlow Inc. becoming Blue Coat Systems Inc. We believe these two companies have worthwhile offerings.
However, there are many more security products than real problems. Keep in mind that at the end of the day, IT in 2003 will be about efficiency and supporting business productivity, of which security is just one leg of the stool.
Businesses fear losing customers enough that theyll pay for software to help retain old customers and gain new ones.
The outlook for business applications, especially SFA and CRM, is bright for next year-at least its brighter than most other categories of software.
The reasons are simple: Businesses need customers, and they fear losing them. And businesses fear losing customers enough that theyll pay for software to help retain old customers and gain new ones.
And, by the way, sales force automation and CRM applications are getting much less expensive (finally). Vendors in these spaces are also shifting focus from large enterprises to the midmarket.
Its amazing CRM vendors didnt try this before. Well, actually, a couple did. Intel and SAP AG set up Pandesic-basically an Intel server running a scaled-down, pre-packaged version of SAPs R3 ERP (enterprise resource management) package-but Pandesic failed miserably.
At the same time, PeopleSoft Inc., Siebel Systems Inc. and J.D. Edwards & Co. were focusing on the midmarket. They are just now beginning to have some success.
The reason it took so long has less to do with software than with midmarket dynamics. Small-to-midsize companies tend to have informal business practices and fluid operations. This is the antithesis of an SAP R3 implementation, which usually attempts to force-fit company operations into a single model.
In the last two years, however, there has been a change: Midsize companies are standardizing on business processes while the SFA, CRM and ERP vendors have added flexibility to their offerings.
Part of that flexibility comes from pricing. PeopleSofts and Siebels midmarket offerings are less expensive, more modular and otherwise scaled back from these companies enterprise offerings. Because of the economy and the competitive landscape, well see enormous price drops in this space next year.
There will also be substantial differences in software next year. For one thing, Microsofts CRM solution will be available, leading the midmarket CRM space to commoditization. Eventually, Microsoft will incorporate technology from its acquisition of Great Plains Software Inc. and Navision S.A. into its software, making Microsoft the leading CRM vendor in terms of clout.
This will lead the rest of the industry to go one of two ways: Either theyll move to a hosted offering, or theyll target vertical markets with their products. At least one company-Best Software Inc., owner of Saleslogix, Act and Peachtree accounting-will move to a completely integrated solution.
Low cost of implementation, minimal administrative costs and a fairly good feature set make hosted CRM and SFA solutions a good fit. Those solutions will appeal to small companies requiring SFA and CRM and to departments in large companies. Those departments, tired of waiting for IT approval, will simply fund the solution out of the department budget and get up and running in stealth mode.
Next year, well see hosted solutions gain more features. Salesforce.com Inc., for example, is adding order/inventory and expense modules to its offerings. Eventually, Salesforce.com will be a complete CRM solution.
In the hosted space, 2003 will see hosted systems acting like extensions of the corporate data centers. Instead of being a detached service, theyll become utilities providing integration capabilities via Simple Object Access Protocol and XML.
Once the integration APIs are fleshed out, well see the first hosted tools. Theyll be simple at first, but eventually hosted solutions will offer full-fledged deployment capabilities. This will blur the difference between hosted and installable packaged applications.
Meanwhile, packaged applications themselves will become services. This trend will start next year with large organizations becoming the CRM service provider to their suppliers and extended sales force.
Of course, for all this to work, integration will remain the biggest issue. On the standards front, things such as electronic business XML will gain steam.
Others will follow, eventually providing an application mesh. 2003 will just be the beginning, but its an important beginning.
For now, the midmarket has other things to worry about.
For one thing, there are few best practices that apply to the midmarket, which means vendors will be adding them. (Theyll be integrated into PeopleSoft next year.) Soon after, the rest of the world will follow. How well those best practices apply to the consumer will be the main differentiators among the vendors products.