Most enterprises that run with a shiny new System i are intimately aware of its virtues. They simply let the system run their core business applications or strategically use the ongoing evolution of the System i to simplify and extend the strength of their companies.
Some System i professionals, though, face new challenges, which can come, for example, in the form of a rookie CIO who, at best, can't spell System i and, at worst, believes that "System i" is just a new name slapped onto an old AS/400.
Other pros fall somewhere into a hazy gray area in which upper management is barely aware of the System i but is bombarded by problems and potential solutions from other platforms. In these situations, ironically, the core strengths of the System i can come to haunt it. If a System i never goes down, runs virtually alone, never draws attention to itself, and the few people in charge of it are always able to leave work at 5 p.m., what happens?
It loses mindshare.
You've probably heard the notion that any press is good press. This principle works for toys, video games, politicians, fad diets, Wintel-based servers, and Hollywood starlets. Does it apply to the System i?
IBM would probably gain more traction if a real disaster occurred, something less mainstream than, for example, a scenario in which a hurricane destroys a building but the System i survives. Perhaps a jealous and disgruntled Windows, Unix, or Linux administrator could figure out that many of his manual daily tasks are inefficient and no longer necessary once a System i enters the picture. Maybe this guy waits until 3 a.m., when even the Windows professionals have left the building, to tiptoe around the System i like Wile E. Coyote, stacking dynamite and slinking out of the data center trailing curly wires and heading to a detonator.
The ensuing explosion would, of course, garner a lot of media coverage — but not because the System i could survive such a blast and keep on running. Nah, the press would have a field day when the perp realized that the System i's off-site, high-availability system simply kicked on and kept the business running. A day or two later, a well-meaning reporter would drive the bad guy insane by telling him that his company was able to turn on more processors with IBM's Capacity on Demand feature and create a bunch of new Unix and Linux partitions, not to mention attach a BladeCenter. Do you see where this is going?
If you have a CEO, CIO, or wily colleague who is not up to speed on the System i, these top 10 reasons why the System i rules will provide a bit of help.
IBM hasn't just renamed the AS/400 over the years — it has technically enhanced it so much that today's System i only resembles its predecessors, yet previous generations sport applications that still run on the newest hardware and operating system. Its own sophistication sometimes bites the reputation of the System i on the behind. Because companies can continue to run "old" business-relevant applications on a savvy platform, the System i itself is sometimes labeled "old" by association. Actually, the System i does it all. It operates and manages multiple application environments on a single server with logical partitioning (LPAR), has handled 64-bit architecture for so long that it's old news, supports up to 64 processors, and plays well with Java, hot technologies, and movements like PHP and service-oriented architecture (SOA).
"We have customers who are still running 5250 workloads in addition to state-of-the-art, cutting-edge WebSphere, Java, and PHP workloads," notes Mark Lencioni, a senior System i product specialist for Sirius Computer Solutions. "You can run really modern applications on the box, and you don't have to throw away any of your older application code . . . you've got investment protection."
The System i's inclusion of IBM's DB2 relational database is its biggest and most noticeable cost-saving "extra," but its other built-in components span security, backup/disaster recovery, database and systems management, communications, networking, WebSphere applications, and virtual Java — all of which, says Elaine Lennox, IBM's vice president of System i marketing, contribute to the proof that the System i installs and maintains programs easier than other systems. "We've preintegrated all the software elements — preconfigured the database, preintegrated the security software, and reintegrated the virtualization and storage virtualization software," Lennox notes. "You could buy all of those components separately and build, ultimately, the same functionality, but that integration is a set of savings right up front — you don't need to employ people to do it for you."
Because the System i comes equipped with key components that work flawlessly together, you don't need a big IT staff, and you don't need to configure security levels and waste time applying patches and updates.
"The System i just doesn't need a lot of attention to run your business 99 percent of the time," says Trevor Perry, IT strategist for Clear Technologies. "The difference is, you can buy a bunch of smaller servers and hire people to manage it. In doing so, it looks like you're doing something. The problem with the i is that it's invisible because it's just working and happy, and that's the huge value proposition most people don't understand.
"The difference, then, [if I were a bad consultant] is that I'm not going to be able to do much work for you on an i; therefore, I'm not going to encourage you to buy an i. And, therefore, I'm going to tell you how the i is a piece of crap and is old and doesn't work because it's more advantageous to me if I sell you an x, because then I can sell you Wintel and Oracle and I'm going to have a lot of work to do, and that's going to keep me in business. That's the enemy we need to put on notice," Perry adds.
You can't blame everything on other operating systems — sometimes business application failures can be traced back to the applications themselves. Fortunately, the System i's long history with ISVs worldwide has given it a powerful legacy of rock-solid business applications that cover all the major industries, such as manufacturing, distribution, banking, and retailing, and include numerous industry-specific solutions as well.
A little more than a year ago, IBM recognized that the System i community was losing ISV application ground and launched the System i "Initiative for Innovation." IBM reports that this program reinvigorated the System i world by delivering more than 2,400 new or enhanced solutions. In addition, because System i sales are closely tied to application purchase decisions, IBM credits the initiative with being a key driver in its addition of 2,500 new clients last year.
The System i is one of the most easily scalable and versatile platforms you can buy. First, a System i 595 calibrates well into IBM's z9 Enterprise Class mainframe territory and even eclipses the mainframe's MIPS processing power ratings when you convert the system's CPW prowess into an MIPS equivalent. There are lots of ways to argue which machine is stronger, but the point remains that if your business grows fast, your System i can grow with it.
Second, the System i is scalable on a simple application level. Instead of running one application per server, the System i can run many applications, and it doesn't get bogged down or confused about which set of instructions it needs to pay attention to in any given millisecond.
Third, with IBM's Capacity on Demand, you can ramp up and cycle back the processing power in your System i whenever you need it. If your business is seasonal, you can call IBM and turn on an additional processor, which is already inside the box waiting for you to wake it up and use it. Capacity on Demand lets you run your System i hotter without having to pay for processing power that you don't use most of the year. If a System i sizzles, it doesn't mean it's going to melt, crash, or break.
In the PC world, many people equate faster chip clock speeds with better performance. The reality is that performance includes so much more — disk and data-access rates, network infrastructure, and the type and amount of memory you use — all of which depend on what you need from your system. If you want to feed several thousand Boy Scouts at their annual jamboree, would you ship the food in a single truck or send a sports car back and forth until everyone has a hot dog?
System i boxes (along with IBM's System z mainframes) can handle more units of work than most other servers, and the System i's performance hinges on how many things you can shove through it at once.
"How hot you can run the box is not just a matter of CPU utilization. It also has to do with the bus structure and the ability to get at I/O. There are a lot of things that go into performance," says Louise Hemond-Wilson, an IT architecture consultant for IBM. "I always tell people, 'Performance is what you do slowest, not what you do fastest.' I could take a Ferrari to the grocery store, but the slow part is doing the shopping. I can get my time to the grocery store really low, but it doesn't have an overall effect on my transactions. If I want to impact my performance, I would attack that big piece, which is plodding through the aisles at the grocery store. The same holds true for computing performance. The CPU is one aspect, and what's good about the POWER architecture is the balance of the different types of architecture that support feeding the processor to keep it hot."
During the last few years, IBM has invested heavily in open standards-based solutions, which give customers unparalleled choices. The System i was the first server in the industry to adopt the Apache HTTP Server Version 2, and the System i group has delivered support and solutions centered around Java, SQL, XML, and Linux as well as gobs of application portability standards and communications protocols.
Wayne Kernochan, president of Infrastructure Associates, says IBM's combination of tight integration keeps in mind many open standards-based applications while at the same time maintains a strong layer of baked-in robustness, giving the System i the best of both open and proprietary worlds. "If there are parts of the server you don't like," he says, "you can bring in other people's software . . . but the underlying architecture is still the integrated stuff that lets you do better [than assemble-it-yourself solutions]."
IBM leaves its mark on the hardware side, too. Recently, the company delivered an industry standard iSCSI card — the iSCSI Host Bus Adapter for System i — which lets you attach IBM's System x Windows servers and BladeCenter, as well as opens the potential for using the adapter with non-IBM solutions.
What all of this really means, of course, is that IBM has systematically delivered a solution that is highly adaptable to technology change. Consider an industry-wide push for standards-based SOA. The System i already plays very well in an SOA world, and anyone who understands IBM knows that as long as customers are interested, the company will continue to use open standards.
Through a combination of technologies, the System i can run LPARs and use IBM's Virtualization Engine to slice and dice system resources for use with specific partitions. In addition to simply being able to run Linux, AIX, and Windows (via an Integrated System x server or direct attached server), the System i integrates it all better than any other server (see "Monitor and Manage Integrated Servers [2]" to learn more).
"It's the server consolidation platform of choice," notes Stan Staszak, director of System i products for Sirius Computer Solutions. "If customers are looking to simplify their environment and virtualize their IT infrastructure, what better box is there to put it on than the System i? You can host more workloads, run different operating systems, use the box like a SAN, take advantage of virtual disk/virtual LAN, shift processing and memory resources back and forth, take advantage of single-level storage for faster disk I/O, increase those server storage spaces on the fly — those are the key advantages and differentiators."
IBM's Hemond-Wilson regularly sees customers who manage Intel servers through the System i's Integrated System x technology, and these customers usually employ fewer people to handle more Intel servers on System i than the companies employ to manage standalone external servers. In one instance, one full-time person could manage 200 Integrated System x servers while another non-System i full-time person could manage only 40 standalone servers.
"To my mind, the System i offers an almost unprecedented methodology for dealing with your IT requirements in a very strategic way," says Charles King, principal analyst for Pund-IT, Inc. "For small businesses in particular, IT tends to grow in an organic way. Your IT operations management comes in and says, 'We're having problems with two servers and have new applications coming, so let's buy a couple of new servers.' If businesses are ready to recognize the benefits of thinking more strategically about IT and are willing to make the effort to make a strategy, that's where the System i can really deliver significant benefits."
Consolidating servers and saving money by more efficiently managing multiple workloads on a single System i are the two biggest ways most customers execute forward-thinking strategies. You can also consider your System i as an SOA backbone for new and existing applications, develop with Java and PHP, use your System i as an easy-to-manage SAN, and even try the upcoming VoIP solutions from 3Com and IBM to save on traditional phone costs and add business advantages at the same time.
"We don't think clients buy the System i because they want to spend less on IT," notes IBM's Lennox. "We actually think the smartest, most innovative clients own this platform because they want to get value out of the investments they make in IT, and paying a squad of people to babysit a room of Intel servers doesn't bring any value to your business."
The System i is the most cost-effective solution for performing business-related work. That fact may seem counterintuitive given the system's relatively high cost of acquisition compared with other servers, but it's really not that surprising. Because IBM includes so many features, applications, and tools, you don't have to pay as much in the kind of upgrade costs that come with other servers to get the same functionality, nor do you have to add another server to the mix every time you want to add another workload. Also, because the system is so easy to install and manage, you don't have to continually reach for your wallet to pay for lots of people to care for it. "It's not the cost of the [competitor's] $2,000 server," Lennox says. "It's the cost of the guy to manage the $2,000 server."
How does the System i keep its checkbook balanced?
That job falls to a little-known cluster of IBM Laboratory Services folks dubbed the Scorpion Group. The Scorpions visit customer IT centers, by request, and calculate the annual operating costs of each platform in an organization. Armed with this information, upper-management people can make strategic decisions based on which systems deliver the most work per operating dollar.
To get hard numbers, the team employs financial and technical experts to ensure that both sides of the fence are properly represented. In addition to applying a rigorous methodology that the group has honed over the years, its members also use proprietary IBM tools to capture and calculate cost and efficiency.
To get measurable data, the Scorpions start with information provided by customers that documents which servers are doing which workloads and how many staff members are allocated to manage those tasks. They also use benchmarked units of work from Ideas International to compare how much is done per platform. Once they have this information, they can calculate how much it will cost to move a job to a particular platform to optimize spending and performance.
IBM collects customer data and then goes on site for two to five days to verify the data and make sure that all the numbers reconcile. The company identifies what each server is doing and determines how many full-time people it takes to manage each server and the amount of work those servers return to the business. The result is an incredibly detailed picture of annual operating costs — and one in which the System i shines.
"So far, in the three and a half years in which the System i has been included in these studies, it has always had the lowest annual operating costs and the lowest cost per unit of work done," IBM's Hemond-Wilson says.
How cool is that?
The System i is the most easy-to-manage combination of enterprise class hardware and operating platforms in the world. For decades, IBM's teams have been working to make it seamless, easier to use, more efficient, and powerful — and they consistently succeed. Sure, the marketing has received mixed reviews, but few customers complain about the technology. In fact, here's a customer anecdote that sums up the simplicity and supremacy inside every System i black box:
"A very large retail client recently told me that it had to do a system migration to a new location," Lennox says. "The company had Wintel, Unix, and System i servers. The CIO called in all the individual departments and said, 'I need you to write me a plan of how long it will take you to migrate to the new building.' So, each department came in, one at a time, and said, 'I'll have to unplug my servers, and then it will take me three weeks to reconfigure them, and I'll need special power and network connectivity, and I'll need this and that. . . .' The fastest a team could get back up and running was in three weeks. Then the team that was responsible for the System i came in and said, 'Well, it takes about an hour and a half to drive to the new building and 20 minutes to plug the system in, so I reckon it'll take maybe two hours.' And the CIO said, 'No, I don't believe you!' and he made a bet with the guy in charge of the System i. I met this guy, and it turns out he absolutely did it. There was a little trouble with traffic, but they still got the system moved and up and running in under two hours."
Chris Maxcer is an industry analyst and contributing editor to System iNEWS. He says his biggest challenge in writing this article was choosing the most important features of the System i and hacking the list down to 10. "There are dozens of cool features that didn't make the cut," Maxcer says. "Still, what's particularly interesting is the strength of the System i in our Top 10 List, which isn't made up of 10 hokey categories. These 10 attributes represent serious business value." You can reach Chris at cmaxcer@systeminetwork.com [3].
|
IBM's Crisp New System i Message
|
| IBM faces three primary marketing challenges with the System i. The first is describing the system to someone who is unfamiliar with it. The second is overcoming myth-based objections such as the contention that it's a legacy system operating on its last legs. The third is communicating that IBM is actively investing in the platform for growth rather than maintenance revenue. Elaine Lennox, IBM's high-energy vice president of System i marketing, says that to combat the lack of knowledge and the misconceptions, IBM needs to help people articulate the story of the System i. "One of our challenges with this has been that we can say a thousand good things about the platform, which we have been doing, but that leaves everyone confused as to what the core value is. We say it's upgradeable, has a 64-bit architecture, is object-oriented [and on and on and on] . . . and we just leave people glazed over," Lennox explains. "Now, we're really going hard on trying to crisp up our message, which is: 'Why do you want to own the System i? Because it will install faster, you will maintain it easier, and you can invest the savings in business innovation." — C.M. |
Links:
[1] http://systeminetwork.com/author/chris-maxcer
[2] http://systeminetwork.com/article.cfm?id=20637
[3] mailto:cmaxcer@systeminetwork.com