I think the facts still count for something in this world. Makes good sense to me, and I hope to you, too.
In that spirit, you can be sure IBM's competition will try to tell you some strange things about System z and the mainframe to convince you they have something better to offer. I would urge you: If you hear negative things about the mainframe from our competition, come back here and post a comment. I can quickly get the facts for you so you can make a sound decision about your IT infrastructure.
As a start, if you're trying to sort out fact from myth, IBM has a Get the Facts web page to communicate some basic truths about its hardware platforms, especially the mainframe. There, you'll probably be interested to read the following, among other points about System z (and I quote):
- "z10 mainframes can outperform HP Superdomes on power consumption by a factor of 2.5x or greater. The key is the capacity of the machine to run work in the industry’s leading virtual and automated enterprise systems environment, not just comparing a single frame z10 to a single frame HP Superdome.
- "IBM's Academic Initiative for System z — creates thousands of mainframe-trained graduates every year. More than 400 universities worldwide have joined the Alliance to teach mainframe and large systems skills - up from just 24 universities in 2004. Over the past three years, nearly 50,000 students have participated in mainframe training and education.
- "HP Superdome configuration was 62% ($11.8 million) more expensive than the [equivalent] System z10 configuration in a three-year TCO comparison."
You can find those facts and more at the IBM Get the Facts web page. Check it our for regular updates. And while you're at it, have a look at the Move Up to IBM System z, also full of facts that set the mainframe apart from the competition.
You get the picture: blinding rain, dangerous flooding, relentless winds that blow down trees and overturn cars. Plus, no electricity, unsanitary water, and survivors on the prowl for dry shelter and food. Does this sound apocalyptic? It should. We are living this reality with increasing intensity provided by global financial crisis.
If you knew the monsoon (tornado, hurricane, tremor—take your pick) were coming, what would you do that was prudent? Gather matches, dry food, blankets, and first-aid kit. Find sturdy shelter in which to hide until the super-storm subsides. And add a couple of other items…radio with fresh batteries, a Swiss Army knife with umpteen tools including screwdriver, file, mini-saw, corkscrew. (Corkscrew?)
Nobody, at least not today, is surprised to learn that IT is severely curtailing purchases of capital equipment. Make that “goods and services.” But, we don’t hear a total cancellation of all purchasing. In fact, Analysts, among others, have adopted a new guidance topic: “The purchases that make sense in today’s Global economy.” Some talk about organizations buying assets that enhance worker collaboration. Others recommend buying only those systems and software that enable better corporate decisions. Still others propose buying what delivers stronger security.
Few have explicitly stated, but all point to the phrase: “Buy only what is essential.” That “essential” is a descriptor which we have typically not, until late 2008, associated with buying any item—whether as personal consumer, or IT decision-maker. For years, most IT orgs bought systems with a variety of motivations. Sometimes those newly acquired platforms supported core business processes. Other times, servers were acquired just because they represented “cool” technology with interesting potential. Today/this minute, the dominant purchase question becomes, “Do we really need ‘this’ or can it wait til our business recovers?”
Which brings us to IBM’s System z. If ever there were a time the word, essential, applied to the new mainframes, it is now. Think this statement overstates the case?
- Is your organization’s intellectual capital as safe as it needs to be?
- Does your customer data ever appear at risk of invasion?
- Do your systems run at full or near-full capacity? Is Utilization where you want it?
- As your workloads spike up and down, do the costs for delivering IT services?
- Can your business sustain a systems outage, especially now?
- Do you want to provision more quickly and according to specific business needs?
- When you think, “IT Investment,” what type of system comes to mind?
- How green is your datacenter; how much energy expense do you now need to save?
- Where does “sprawl” fit into your portrait of your IT resources/physical space?
- What type of single system do you need to execute multiple, mixed workloads?
- When did you last closely consider today’s mainframe for cost-savings?
Obviously, every answer to these 11 is intended to point to IBM System z. Why? Because the word, “essential” applies to systems that, at minimum, enable service levels to be met, if not exceeded—at the lowest possible cost. Nobody needs risk, especially now. Nobody wants to invest in resources with short shelf-lives or whose architectures are not forwards and backwards compatible. Essential equates with “must have,” versus “nice to have.” Essential applies to the foundation, the core, the nucleus of IT operations.
This COMMON SENSE version is not intended as a lecture on prudence in a storm….or is it? What we want you to understand—and discuss with us if you don’t—is how IBM System z10, both Enterprise Class (EC) and Business Class (BC) more than meet the criteria that firmly establish an IT resource as essential.
Super-stability of System z saves on outage costs, helps retain customers and reinforces service-level predictability. Government Security certifications reinforce z’s risk management leadership in this vital category. Who can absorb the shock of stolen confidential information, and exposed military plans and defense systems? When could systems unplanned (and even planned) outages be worse to maintaining business continuity?
Fear-mongering runs rampant on Wall Street and other global financial centers. Fear generates irrational decisions about what to buy, what to sell. System z was built to overcome concerns about the ability of any system to perform to customer requirements…including dynamic needs fueled by a variety of dangerous circumstances. With an MTBF design point of greater than three decades, you can expect System z to exceed your availability expectations. With Gartner Group ratings that put System z at the top of Consolidation platforms, you can expect System z to take sprawl and put it in its place.
Bottom line: we can juxtaposition Global economics and their implications with System z until we are tired of keyboarding…or you stop reading. We prefer that you prepare a list of your “must haves” for the next 18-24 months of your IT operations. Then, let us meet with you, go over your list, and provide factual evidence of how System z can be essential to the well-being of your IT—especially during a Monsoon.
System z’s reputation was not built on fabricated stories, false claims, or price-gouging. We have proven we are essential to the major financial institutions, government agencies and other industry sectors. With the z10BC version, we are now essential to small-medium business.
Just one example of savings and value: The IBM System z10 Business Class has the capacity of up to 232 x86 servers (1) with an 83% smaller footprint (2), up to 93% lower energy costs(3), and a much higher level of security, control and automation – allowing for up to 100% utilization.
Let’s talk now. We welcome your toughest questions and support your concerns.
- Source: all performance information was determined by IBM in a controlled environment. Actual results may vary.
- Source: 1 quad-core z10 BC running 10 IFL's takes up 30 square feet. Compared to 232 single core x86 rack servers which takes up 150 square feet.
- Source: 1 quad-core z10 BC with 2 I/O drawers running 10 IFL's uses 4.5 killowats of power. Compared to 232 single core x86 servers which use 67.0 killowats of power.
Tens of billions of dollars—no exaggeration--are being lost or forfeited due to security breaches, worldwide, in 2008. Private citizens, national and local governments, and global enterprises fall prey to this BIG sting. Medical and banking records are exposed. Intellectual capital loses its privacy. Identity theft has become “common.” News reports across the globe say various governments use cyberspace assaults to access another nation’s key defense information. It does not matter which country; exposure of critical assets threatens us all.
When we “feel” insecure, some of us strap on seat-belts, don helmets, double-lock our front doors, install security software, change passwords. We encrypt, we hide, and sometimes camouflage our assets, whether personal or corporate. Simply, we take defensive measures to try and ensure that our risks—and fears about them—do not materialize. Most times we are successful…most times. Those who believe that personal, governmental or corporate assets are completely secure may be seriously out of touch with reality.
The 9/11 events jolted business and government into tighter security…or the illusion of it. IT had either a field-day or nightmare, depending on one’s technology function at the time: conduct security self-audits; source new security software; consider industry analysts’ recommendations on improving baseline IT security. Today, some in IT think security has been solved by software and middleware…layers of it….big stacks of it. To some extent, this perception holds true.
Can IBM mainframes prevent all security intrusions? Perhaps. Who knows? The most important discussion points may be: how integral security is to core service levels; and, how it can consistently safeguard data, records, and intellectual capital, many companies’ source of competitive advantage. Our pitch? Consider more closely IBM’s System z for overall security, and as the security “hub” should you plan to implement an SOA environment.
In an ideal world of enterprise computing, those heading IT security would put an umbrella over all resources in the complex: servers, storage, hubs, routers, switches, and sub-systems in between. Little would be left to chance. This umbrella would address known hacker techniques, histories of “invasions,” and past exposure to information assets by those who should have no access. All relevant government certifications would apply to the entire operating environment. But the question remains: is any security literally “impregnable?
Data proliferation, new governmental regulations, and increasingly virtualized distributed environments have made security—how to establish, maintain, and continuously improve it—paramount. Fortunately, IBM mainframes are designed with security as a primary focus. That statement applies to how the server, operating systems, and middleware are architected. Bullet-proof is the goal…and, relative to alternatives, the achieved objective.
When Butler Group (analyst firm) says that 70% of critical applications in Government run on mainframes (1), reliability, availability and scalability are not the only reasons. When the combo of EAL5, FIPS 140-2 Level 4 and related security certifications are earned by System z—and only System z—the other key reason becomes clear. Exciting? Hardly--but essential to any company’s ability to execute against business goals without undue concern about the security of their vital information, and speed by which that vital information can be communicated.
Security expertise, alone, may not provide a sufficient cause to shift your systems to IBM System z. But, there are questions worth asking either to affirm the current IT operating environment, or to consider a change or makeover:
- When was the most recent compromise of corporate IT security for whatever cause?
- What was the business impact of that breach, and how was it remedied?
- Have the Lines of Business you support stated their security requirements?
- Does your “Security Committee” constantly update requirements?
- Is the security software you use with your hardware causing performance hits?
- What is the priority your company places on data center security?
- Do you perceive security ranks in the top 5, top 10 IT improvements you want?
- What would your business be able to accomplish if your IT were more secure?
Note 1 -- Butler Group’s Roy Illsley: ftp://ftp.software.ibm.com/software/systemz/pdf/analyst/The_King_is_Dead_-_Long_Live_the_Mainframe.pdf
Most IT organizations will not flinch when answering. Many will even re-affirm their operating protocols for security. What some may lack, however, is deep understanding of IBM mainframe security advantages. Consider the following:
- Mainframes execute multiple, mixed workloads with remarkable integrity and integration. This type of tight integration between images on a mainframe -- z/OS, Linux, and z/VSE hosted in dynamic virtual partitions -- provides outstanding network security. Example: the design of a System z innovation, HiperSockets, provides an internal network connection between virtual images on System z. HiperSockets can protect the data communications that distributed networks often expose using TCP-IP. Fortunately, again, System z has secure methods for working with distributed systems and encrypted networking.
- measures systems performance, especially that of mainframes. A single IBM System z10 Crypto Express (1) card can enable up to 6000 SSL handshakes/second Implication? System z is outstanding for secure, high-volume Internet communications: data serving, web access serving, file transfer, email and more.
- Encryption provides another level of protection from “sniffers” and “snoopers,” helping to ensure that only the intended party is allowed to decrypt sensitive data. System z provides encryption solutions to protect data at rest and in flight. In fact, IBM encryption acceleration can be provided with every System z, along with other outstanding, hw and sw security features that are designed to be tightly integrated. You get advanced security with minimum throughput penalty!
- Business Continuity (BC) and Disaster Recovery (DR) have rarely, if ever, been more in demand. In response, organizations rely on networked, geographically dispersed facilities, and on virtualization and other tactics to preserve service level delivery. Problem is, each of these “techniques” can compromise security. Test System z, again, to determine how it provides BC and DR without the same compromises.
The goal of this entry was to whet your appetite to learn more about System z security. We invite you to connect your security experts with ours. Our story is much deeper and proven than what you have just read. And if you do not yet believe that security is still a very live issue, pop open these links: You may have second thoughts.
PLUS, read “our story:”
Note 1 -- System z is the only server with EAL 5 certification for logical partitioning. Operating systems certifications include z/OS at EAL4+ and Linux on z at EAL 4+.
Talk about mixed messages!
Notice how vendors position features of their distributed systems by terming them, “Mainframe-like.” Then, they publish statements that disparage the Mainframe. Which is true? Do competitors really design their new systems to emulate IBM System z’s best attributes? Or, do they consider the Mainframe to be a relic, an artifact of the beginnings—but not future--of enterprise computing?
This installment in COMMON SENSE aims to stimulate thinking so that your next server purchase, for enterprise-level tasks, hits the sweetspot where IT and Business converge. It’s sometimes challenging to determine if you want the genuine item, a Mainframe, or some “imitation” whose appeal is; “just add another cheap one of these if you run out of capacity.”
Hundreds of articles have been written—and, no doubt, you have read at least a few—about the Mainframe. We trace its origins to around 1964. What is important is that Mainframes began decades ago…and have persevered; modernized, and delivered a rich set of benefits, clear through the moment you read this. With recent IBM System z10 launches, “the Mainframe” got another renewal….one showcasing unprecedented cost-value driven by: enhanced capacity, throughput, security, RAS, and virtualization.…Enough bragging for now.
Back to Mainframe-like. Why do System z’s competitors disparage the Mainframe, then proclaim that their enterprise platform is just like one? Consider parallels to this behavior in other industries. Examples: saying a hybrid car’s acceleration is Ferrari-like; or, promising “Movie-star-like smiles” with an over-the-counter tooth whitener. Does this approach to creating market appeal really help IT decision-makers make the best possible decision about which asset—with most business value--to deploy?
Whether we just asked a rhetorical question or not (we probably did!) is less important than penetrating the marketing fluff and getting down to what system delivers the best for an IT buyer’s pain points. So, let’s quickly profile what sets apart a Mainframe from Mainframe-faint replicas.
- Mainframes capably execute the largest, most complex mixed workloads. Think of your own IT requirements and ask how many servers you use to handle tasks that access hundreds of millions of records, for example, in the same day…or even hour.
- Mainframes consistently demonstrate that they can function at sustained 90% utilization…or more. Utilization varies by architecture. Mainframes, because of their core architecture, should be run hard without undue concern about failure, especially with System z’s MTBF design of greater than three decades.
- Mainframes virtualize complex enterprise environments with unparalleled granularity. Why that’s important denotes the financial necessity of right-sizing workloads to resources. When that happens, task-execution can take place at the lowest cost per transaction.
- Mainframes can execute tasks with incredibly low use of energy per transaction. System z has pioneered “green” in many aspects including: how it cools, how it manages energy use, how it conserves. Ample proof exists to support this statement.
- Mainframes achieve the highest security certifications like EAL 5 and FIPS-4. With the countless attacks on data and intellectual property, only the highest security makes sense. There is no middle ground when privacy and $Billion’s are at risk.
- Mainframes can absorb big, sudden activity spikes without significant scale up of the people supporting them. Mainframes were designed for dynamic businesses. Mainframes distinguish themselves with their automated, on demand capabilities.
- Mainframes provide remarkable availability across vast geographic distances. Global business means globally connected systems, which must collaborate to meet or exceed aggressive service levels. Mainframes, using Geographically Dispersed Parallel Sysplex (GDPS) are designed precisely to do this.
- Mainframes handle more transactions--more quickly--than any alternative. A look at core banking benchmarks (such as Bank of China) attests to this. As premier data and transaction hubs for SOA, Mainframes help ensure astonishing throughput.
- Mainframes provide both flexibility AND control that help ensure Service Level Agreements (SLA) are met. SLAs now include security and disaster recovery metrics. These indicators of system value to business continuity point to Mainframes.
- Mainframes handle enterprise-size workloads within a refrigerator-sized footprint. System z’s early ancestry required rooms, raised floors. No longer. Today’s Mainframe economizes space and delivers stupendous performance in that footprint.
So, there are Mainframe “imitators.” That’s ultimately the most sincere form of praise we believe. It just strikes us as ironic that those same imitators try to profile System z as “the past.” Reality says that essentially all the mission-critical server technologies in use today, whether in the RISC or x86 space, were inspired by the Mainframe! Simply, today’s Mainframe represents the past and provides the innovation—now and beyond—that assures both its own future and that of “Mainframe-like” offspring.
Today: data and regulatory proliferation, data and privacy invasions, space shortage, power availability, and escalating business requirements demand systems that come closest to the “ideal.” Maybe you will buy the “Mainframe-like” alternative, or continue using the one you may have. But, let us show you proof of our attributes, and we believe you will determine that you need the original, not the imitation.
This COMMON SENSE installment started by challenging how Mainframe alternatives can confuse buyers by terming certain system features as “Mainframe-like” or, “Mainframe-inspired.” The real story seems to be this: buy the system that time has tested, that consistently innovates, executes, helps provide long-term investment protection, and can seamlessly adapt to the dynamics of Business Year 2008 and way beyond. Let’s, please, continue this important discussion.
These are tales about the new realities of the mainframe. I'm trying to spin these yarns without spin. Just to attempt to bolster my own credibility to do that, I've given this entry a subtitle stolen from that great Miles Davis tune. You might ask yourself that question somewhere along the way as you read this (as I do), but I am pretty sure that by the time I get through this whole story you'll see why this matters to you. Anyway, here goes:
In a perfect IT world, the Lines of Business would routinely volunteer that IT makes their business hum. And data center staff could say that their resources all work effectively: none provides a security risk, all are “Green,” and they do not waste precious real-estate. The IT architects would say that the data center structure can support any application or process that their business requires to compete and win. And the operations people would tout the service levels they could guarantee, at a cost the CFO praises.
Hmm....IBM just announced that getting to their perfect world requires an IT makeover. Those 3900 servers that have been “running” IBM are now shrinking to around 30, all Mainframes, and running Linux® on z/VM®. Virtual Linux servers running on z/VM can scale for small-andmediumsized businesses as well as large enterprises. z/VM’s “gold standard” of virtualization helps enable IT users to derive the most economic value from assets, AND helps ensure efficiencies by autonomically right-sizing resources to workloads. IBM projects exponentialsavings in footprint, energy use and investment—criteria used by professionals, anywhere, who decide to end sprawl rather than strugglewith it.
You are reading this pitch—-which is exactly what this is—because we hope you consider our mainframe running Linux within our z/VM operatingenvironment. We sense that you have resisted buying a mainframefor reasons that we cannot challenge. The IT world “went distributed” at a time when mainframes were closed, proprietary devices with limited function, thus value. That world has changed.
With the IBM System z™ platform family, we have increasingly embraced and championed what IT professionals have said makes sense: open systems that work well with either specific or de facto industry standards. System z “majors” in Linux and Java™. System z majors in virtualization that helps maximize the utilization of all system resources: CPU, memory, I/O devices, networking, crypto processors, and more. System z works with SCSI devices and tape media from a variety of vendors.
OK, maybe “so what?” to this point. Linux is Linux, whether IBM or HP (or Sun) resells it. Intel® servers, not as expensive (per unit) as Mainframes, can be bought and quickly installed to help cover capacity gaps. Virtualization software like VMware—which we resell with our Intel-based servers—helps improve utilization, even (to a certain degree) provisioning. Have we made a compelling case for System z yet?
Where the mainframe likely captures your interest depends on answers to a few fundamental questions...which, by the way, we asked ourselves before embarking on our own “makeover:”
- Do you have what could be termed, “sprawl,” and it is sustainable,but becoming less so with each new application or service level requirement?
- If you are using the abilities of VMware to control some of that sprawl, do you foresee potential security issues? Have you already experienced any/some?
- Has your virtualization, whether delivered through HP’s VSE, Sun’s methodology, or VMWare created more cycles, absorbed more IT budget/personnel to manage?
- Does Sun create confidence that their roadmap and your hardware asset requirements coincide? What is your discomfort level moving to ROCK or the next Intel or AMD architecture?
- Have you recently talked with those holding your budget about the need to trim expenses while delivering better service? Were concerns cost-only, or investment protection, too?
- For new apps: could you develop, test, and run for production various applications on the same physical machine? How many of each could you do without enlarging footprint?
- Do you anticipate a time in the next three years where you could run out of space or adequate electrical power?
- Has one or more lines of business asked you for help with ERP or CRM in terms of better service levels to support a more competitive business?
No doubt, you would answer a few strong “yes’s and no’s,” and some “maybe’s.” So, let’s talk a bit more System z here. What we most want to convey is this: System z is far more affordable than when you last considered a mainframe. Many people wanted “one,” but could not justify either the acquisition or overall ownership costs. Next, System z and Linux provide the rare combination of economics, security and qualities of service for which mainframes are legendary.
Next, skills to run mainframes, especially for net-new mainframe customers, are proliferating: IBM’s System z Academic Initiative is generating thousands of University grads, worldwide, equipped to care for System z’s. Our own “migrate to the mainframe” consortium, Destination z, now has dozens of key members. This organization affirms our ISV, SI, BP, and Academic “reach.” The Mainframe Charter assures that we will continuously improve System z’s relevance,ease of use, applications availability, overall value and economics.
Still not sure a deeper discussion of the technologies is worth your organization’stime? Consider these recent case studies where Sun was among the incumbent platforms, and System z became the destination through a risk-managed migration and TCO outcome that justified the switch:
- Case study example in US: Erie 1 BOCES builds a powerful portal with IBM WebSphere, System z and Linux “With WebSphere on the System z platform, Erie 1 BOCES has a highly automated, scalable and cost-effective solution which can support growth and promote development, providing a higher level of service to our customers.”—Carol Troskosky, CIO, Erie 1 BOCES. (26 Sep 2006)
- Case study example in EMEA: Nexxar Group transfers to IBM System z9™ Business Class for cost savings “IBM brought together the right people to assist us in introducing an IT Service Management framework, and we are now seeing clear benefits in terms of cost reduction and greater flexibility.”—Wim De Ridder, CIO, Nexxar Group, Inc. (13 Nov 2006)
- Case study example in AP: Aiming for a speedy, simple integrated system to enable economical use of IT resources, Lawson becomes an On Demand Business
- Example showing migration to System z from Sun Systems, where a government org consolidated more than 150 Sun Microsystems UNIX® platform-based Oracle database servers onto a single IBM System z9 109 server running the SUSE Linux Enterprise Server (SLES) operating system under the IBM z/VM V5.2. Configured with five Integrated Facilities for Linux (IFLs)— central processors that are dedicated to Linux workloads—and 48 GB of memory
Though IT professionals rarely talk about a product feature or piece of software as “our special ingredient,” z/VM is it for IBM System z9’s unprecedented Linux value. Yes, Linux can run directly on the real hardware, and as with z/VM, additional operating systems can run without destructive interference, in partitions that all belong to the same system. Where z/VM creates such value is its flexibility and control over the entire environment. Provisioning, security, load-balancing, sequencing, micro-partitioning, tapping reserve resources for sudden spikes….these are a partial listing of z/VM’s contributions to Linux on a Mainframe. Additionally: z/VM enables users to achieve very high levels of resource commitment to help reduce costs and maximize operational efficiencies.
At your request, we can provide detailed metrics about the size and number of Linux partitions, and what are documented utilization rates under specific workloads…..and what are real-world energy savings from a z/VM environment. We can show an MTBF design of 60 years for our mainframes. We can bring to you proof points and testimonials that vividly show z/VM’s impact on consolidation, security, and availability, especially in a Linux environment.
Right now, we want you to consider taking a deeper dive with us, and having more of your technical experts engage with ours around the possibility of transforming your infrastructure to deliver ever-increasing business benefit. Linux and z/VM on System z: arguably the best of both worlds. For more information on what you have just read, see http://www-03.ibm.com/systems/z/os/linux/ and http://www.vm.ibm.com/.[Read More]
Questions to ask:
- Is there such a thing as a ‘good deal’ in the IT world?
- Do IT decision-makers leave their ‘consumer’ hats at home when they evaluate systems?
Let’s briefly look at answering these fundamentally simple, but necessary questions...
Consider the following scenario that would, in most cases, depress us as consumers. It’s a situation that could change buying behavior. You purchase an automobile at a specific price, and then, a week later, discover that same car model went on sale. Why? Because Consumer Reports stated that, through years of use, the overall costs to drive that vehicle proved painfully high. Worse yet, imagine you also find out from Consumer Reports that if you’d bought the more expensive car--the one you really wanted--it would have proven more cost effective to drive in the long run. Buyer’s remorse would set in faster than new car depreciation…….
So, for the next seven minutes, please put on your consumer hat, even if it sits on top of the one you wear when making systems purchase decisions. Keep in mind that value is often relative and can be difficult to determine.
Everyone makes “claims” in order to sell systems. Some draw attention with prices. Total Cost of Acquisition (TCA), for example, is frequently used by x86 and Itanium system vendors because “Marketing 101” teaches that price sells in a commoditized market. Further, this selling behavior is reinforced by “It’s good enough” buyer expectations . At the other end of the spectrum—the high end of the systems market—the sales language is markedly different. The IBM System z™ is best represented in terms of Total Cost of Ownership (TCO). But don’t let these terms and the accompanying proof points become stumbling blocks when assessing system economics.
This edition of COMMON SENSE has two objectives. First, we want to reset your grasp of System z TCO. Second, we hope this brief summary prompts you to request a System z deep dive with your IT operations team, finance squad, and those in IBM who will objectively model how our mainframe might be the best economic IT alternative you’ve seen in a long time.
Keep in mind, there is no absolute “good deal” in IT—or anywhere else for that matter. A “good deal” is a relative description of what somebody initially paid for an asset, and the value that asset delivered throughout its lifecycle. In other words, value can sometimes be a matter of perception; and one that reveals itself over time, particularly when evaluating large systems.
This is why we strongly advocate systematically mapping investment value over time. You see, mainframes pay ongoing “dividends.” With our software models, the more you buy, the more your cost-per-transaction ratio usually improves. Mainframes can seamlessly scale workloads without typically adding people resources to manage the increased volume. Mainframes deliver security levels that can protect your most vital organizational assets: customer data, intellectual property, vital statistics, etc.
These capabilities are fundamental to the mainframe design. And, yes, they do boost the acquisition price; but, most frequently (and importantly) they lower the TCO. With a mean time between failure (MTBF) design of greater than three decades, this TCO advantage can prove and re-prove itself year after year after year.
“You get what you pay for …” We have all heard that. And it’s something to remember when your company’s security is at risk. Consequently, you employ proven security tools which could consist of software, whether homegrown or third party. Add that software, by layers, and you might find your assets protected. Your data transmission rates, however, could diminish. Trade-offs!
System z does not, typically, require those tradeoffs. Its design makes security largely intrinsic—embedded in architecture, components, and subsystems. And, service-level-destroying latencies are essentially overcome by our embedded encryption accelerators. Yes, System z might mean a higher price tag on day one, but if you consider the staggering business costs of lost, stolen, or corrupted information, you will again be thinking; “You get what you pay for …” Or consider the potential damage to customer retention if sensitive data is delivered s-l-o-w-l-y due to system drag imposed by multiple software security layers.
There are no absolute right or wrong answers here either; just different perspectives. Why mainframes may appear to cost more than other types of systems is, again, a matter of context and weighing the overall, long-term costs. We would, for example, refer to the cost savings from decades of mainframe virtualization innovation and its direct benefit to utilization rates. Many analysts, users, and technical academicians consider the depth of mainframe virtualization to be THE standard—and where competitive technology usually comes up short. With IBM mainframe virtualization you can, more cost-effectively, match resources to workloads. And that, ultimately, yields better business results.
The business discussion of weighing TCA versus TCO will persist until a worldwide-accepted methodology presents itself. Many in IT consider IBM Scorpion studies to be objective and empirically valid. (These can be comprehensive eight-to-10-week on-site TCO client evals that link business value with IT. Scorpion studies examine current and alternative IT systems, then apply realistic financial analysis). In well over 1,000 Scorpion studies, the mainframe TCO consistently shows significant advantages over alternative system solutions.
To get a better perspective on all of this, you really need to invite a System z specialist to talk with those on your IT staff who have expertise in security, virtualization, mixed workload execution, and reliability/availability/serviceability (RAS). Be sure your IT staff is accompanied by those on your team who monitor costs of ownership. The discussion will prove very telling as long as costs like; purchase price, test, installation, maintenance, upgrades, software licensing, operations, and outages are assessed. If you do objectively look at the bigger picture, we are confident that we, IBM System z, make the best deal in a world where “good deal” is relative.
Net of this COMMON SENSE edition: We want you as a valued client who—because of your System z-enabled business benefits—becomes a powerful advocate. So let’s work together and determine the ultimate value from buying an IBM System z—versus whatever else you are evaluating to help solve the pains of; server sprawl, sub-par security, uneven throughput, and avoidable complexity. And remember your “consumer” hat when you evaluate IT—don’t leave home without it—especially not in today’s economic storm.
Despite growing defamation by our competitors, IBM’s Mainframe thrives. No matter what half-truths you see in alleged “white papers” by those wanting System z to disappear, we keep generating results that show just the opposite.
Why these groundless attacks on System z? For starters, our competitors do not want to accept that our System z servers appeal so broadly and deliver tremendous value. Competitors falter when trying, factually, to compare their ability to drive down cost of ownership. They cannot match our overall throughput on an apples-to-apples comparison. And they cannot match our highest levels of security certification. But despite this, they will still try to tell a global IT world, facing enormous budget challenges—many imposed by the business of their own companies—that adding to costly sprawl is OK, or that a certain number of security breaches is tolerable.
The System z design center has been/is/will be that complex mixed workloads demand impeccable stability, throughput, security, and control. These tenets guide our development, reflect in our messages to the market, and reveal—most importantly—in the overall user experience. Our goal is to deliver cost-effective, innovative, super-stable systems that enable business. Period! We do not think that mainframes are the only system needed in IT today. But we do strongly advocate that when power consumption, floor space, scalability, business continuity, and overall value are paramount, THEN “it’s mainframe time.”
Weeks ago a competitor published an attempt at a “white paper.” We say, “attempt” because we have criteria for authentic white papers that mean: facts matter, logic underpins any findings, and even skeptics can support the paper’s conclusions. What our competitor wrote brought up the shopworn criticisms. “They” said the mainframe has not enough skilled personnel to support these systems. They said System z’s Linux is really just for web-edge and I/O workloads. They said System z imposes too many unique (read “proprietary”) routines.
Let’s be blunt: there are clear and documentable reasons why System z’s share is growing, and why current users add new workloads in addition to more capacity for existing tasks. There are clear reasons we win among the most prestigious environmental awards. There is ample evidence that worldwide mainframe skills are proliferating, not contracting. Analysts assert this; so do Press. So do academic institutions. So do satisfied users whose case studies attest to consolidation outcomes and net, overall cost-savings that can justify an arguably higher initial purchase price.
Forty-nine thousand university students have completed mainframe courses just in the past four years. Four hundred fifty universities and colleges around the world belong to IBM’s heralded Academic Initiative. Market share for System z, in the greater than $250K segment, has nearly doubled since year 2000. Competitor share continues to decline according to analyst-provided metrics. $100M is being spent on System z simplification. Linux specialty engine growth from 1H-‘07 thru 1H-‘08 has risen 26%, an indicator of worldwide IT confidence in running Linux on System z. Such customer acceptance is fueled by all of our specialty engines. They help absorb software costs while off-loading focused routines from z central processors.
And the z-supporting facts go on, and on, and on. It can take four HP SuperDomes to match the TPS performance of a single z10. At the same time, for the same workload execution, SuperDomes can use over triple the energy, over 2.5x the floor space, and many more people to manage these systems.
What are we getting at here? Simply, that facts support better decisions. IBM System z commits to helping clients buy z only when it makes most customer sense to do so. We care about our credibility and your company’s progress. The competition seems intent on falsifying System z’s demonstrable value. We find that competitive tactic especially inappropriate in a worldwide economic climate where buyers need better information to make balanced, prudent decisions.
So, ask our representatives any questions you have about System z’s comprehensive excellence. Visit our various worldwide customer centers to simulate your workloads. Make independent judgments about whether we bring the most flexibility and control to a global IT environment best described as chaotic and complex. We welcome your close inspection. This applies to whether you own a mainframe and are considering adding new workloads, OR, you have all distributed resources and need to understand if/why consolidation would be best on a System z.
Meantime, skim these statements by Clients, Analysts, and ISV’s. We think you will conclude that System z really is essential, especially in today’s financially challenged IT world.
- "We've implemented z10 systems in our parallel sysplex environment which are running live, mission critical benefit applications for the world’s top employers -- so availability, performance, and reliability are of utmost importance," said Sandee Kotowski, Manager of Mainframe Infrastructure, Hewitt Associates. "The IBM mainframe has been a key part of our IT infrastructure over the years, with clear cost benefits, but this new system takes that value proposition a leap ahead. The capacity and scale of this system changes the economics of the mainframe and is a significant step forward in addressing our constantly evolving technology needs." Sandee Kotowski, Manager of Mainframe Infrastructure, Hewitt Associates
- JungHoon Sung, Manager of IT System Planning of Industrial Bank of Korea (IBK), said, “The success of IBK's System z10 migration was based on our great partnership with IBM in Korea which helped secure mandatory capacity for mainframe, reduce TCO and improve system stability, efficiency, and liability. We expect innovation in client service through the great capability of IBM System z10.” Source: October 2, 2008: http://www-03.ibm.com/press/us/en/pressrelease/25311.wss
- “A growing number of customers are integrating System z virtual Linux servers into their Oracle Grid strategy. IBM’s new System z10BC server extends the reach and the affordability of this solution. The z10BC platform is an attractive option for customers moving to virtualized environments.” Matt Puccini, Senior Director, Oracle Corp
- "We were amazed with the power and flexibility of our new z10. Compared to the cores on our test Intel server, only half as many IFLs were needed to host the Live version of our massive game Taikodom, surpassing even the most optimistic estimates of our Lead Programmer. Web services were also 50% faster on average. Now we have plenty of room to grow as the player base expands, simply by adding more IFLs to the production environment. Meanwhile, those same IFLs can be used for tests and development, and because we are also offloading an increasing number of math intensive operations to our Cell/B.E. processors, there is no foreseeable limit to the number of simultaneous players our z10 will be able to handle. The z10 also proved itself very easy to install, and it's Java compatibility worked better than expected, making it possible for us to move the whole Taikodom environment - including it's web portal and Bitverse middleware - to it's new server in only 12 hours, with downtime limited to 3 hours." Tarquinio "Tarq" Teles, CEO, Hoplon Infotainment SA
And listen to Forrester Analyst Brad Day’s taped interview at www.on24.com/clients/ibm/117759.
KISS, a potentially insulting command, carries risk. Why? First, because hearing “Keep It Simple Stupid” often makes us fixate on the word, “Stupid.” Second, because life's complexity is escalating. Third, because the costs of simplifying sometimes outweigh living with complexity. Technology provides a vivid example. Virtualization, for many, helps drive efficiencies in resource use. But it also yields complexity: added needs for management, security, and licensing. Not all bad, just not “simple.”
Data and application proliferation, global financial hardship, increased regulations, mounting piracy of physical and IT assets—in combination or even separately—point to added complexity, which can drive higher costs and a greater sense of frustration and vulnerability.
IT Simplification...a major quest and need. Make fewer things do more work at less cost. That's an appealing requirement or destination. But it is not one that is easily achievable. The goals of simplification are not always clear. Prioritizing simplification often proves challenging. How many times do we say, “I went for ‘simple’ and got very little in return?” For example, there are some who wonder if going from Command Line Interface (CLI) to Graphical User Interface (GUI) ultimately delivers the best payback. Or if reducing choices of UNIX® operating systems provides a superior user experience. Simplification can be a matter of perspective.
In the ideal World of IT, basic methods for the most cost-effective execution of workloads dominate design. Systems Administrators (SAs), their managers, the IT architects, and apps developers would speak the same language about needs, intersection, collaboration, and desired outcomes. They would also reduce the number of devices, the instructions governing their use, and the rules for device connectivity. The end result: well, some could say the mainframe is the paradigm for simplification. And, many others would shake their heads sideways. What’s the truth here? Is there an inherent contradiction to Simplification?
IBM's reputation for server excellence, from small capacity to immense, began with the System/360™ mainframe family. They were fed and cared for by a handful of highly focused people in white coats, exercising their skills in raised floor, glass-enclosed rooms with regulated climates. Hmmm...somehow that picture does not convey simplicity. In fact, mainframes were (essentially) the only architectures capable of handling large-scale, mission-critical tasks...but still required special language, skills, tools and rules. Simplicity was not the primary design center here: automating and performing large transactions and calculations were the super-ordinate goals; complexity was subservient to need.
It's now 2009. The server architectures have shrunk to a few, along with just a handful of mainstream operating systems: Linux®, UNIX, Microsoft® Windows® and mainframe. Does this equal simplification? If it does, is the IT user community better off than when UNIX had many personalities, OS/360 derivatives mutated, and there was no Linux as a potential unification point? Philosophical queries, pretty much, to this point.
So let's get down to the basics...
- What is your IT organizations first goal for simplification; what are the first measures of success?
- Who is the force behind simplification, and what does he/she expect to see by way of payback?
- What in your IT asset base is untouchable, because “simplifying it” might prove too disruptive?
- Where does your infrastructure today seem most complex and how did “it” get that way?
- How much could you save if you simplified, for example, just your servers (vs apps, networking, processes and storage)?
- Is simplification, like “continuous improvement,” a moving target?
- If you could shrink 3,300 distributed servers to 33 mainframes running Linux, would you consider simplification a “mission accomplished?”
- When might acquiring a mainframe or committing more workloads to ones you have be an act of simplification?
- In IT, is anything really simple any more and how much complexity will you tolerate from a cost/value standpoint?
- When did you last consider data management, apps delivery and security preservation “simple?” Do you measure that consideration in days, months, or years?
No simple answers, we suspect. Except, and a big “except,” we are completing a multi-year $100,000,000 investment solely in mainframe simplification; everything from GUI, to I/O, to programming, to script customization, to automated business-goal-driven IT resource allocation. Simplification, to IBM System z® means a superior user experience from the standpoint of workload execution, business continuity, ownership costs, and skill development. We chose this path to make a clear statement: we will always make our mainframes simpler to understand, acquire, use and maintain, to satisfy sharper market requirements for “faster, better, cheaper.”
Mainframes are not for beginners. That could be viewed, arguably, as a self-condemning statement, especially when made by a System z marketer or salesperson. However, there’s a coaching from Business Development that applies to mainframe simplification assessment: “Go Slow to Go Fast.” Translated: once the mainframe routines are mastered and protocols for effective service delivery become habit, mainframes may be the simplest of all systems today.
Consider this: Bank of China Core Banking Benchmark revealed IBM System z9® mainframes, working together, generating 9,400 transactions per second, while servicing 380 million customer accounts and approximately 1.3 billion account records. Those results exceed by four times what HP claims in its Temenos Core Banking benchmark. And, System z9 mainframes doing that work required roughly ¼ of the energy to execute. What does “Simplification” have to do with these comparisons?
Simple! (Sorry.) What a benchmark like this shows is that once a mainframe is “mastered,” it can execute more work at lower transactional costs than alternatives. So it would seem simplification is both process and result. To a System z novice, operating “the machine” may seem daunting. To the more than 50,000 students at 450+ universities worldwide who have recently completed mainframe curricula, operating a mainframe just got simpler.
To the mainframe owner who just adds specialty engines to the inside of an existing footprint to gain capacity and avoid more sprawl, delivering business enablement just got simpler. To the CxO concerned about sprawl and energy/space conservation, buying a mainframe makes IT’s value-add simpler. To the Line of Business executive using customized scripts from IT to retain customers, mainframes just make life simpler. So, the next time someone tells you to KISS, ask what he/she means. If they seriously believe Simplification is “easy,” remind them that getting to “simple” challenges experts. It is the destination that counts.
Want more proof? Talk, now, with a System z specialist. Ask how the complexity and expense of “distributed sprawl” can be cost-effectively overcome. Let us demonstrate, with validated facts, how System z security simplifies the protection of assets, a corporate urgency. Press that IBM System z specialist and client rep on the automation (simplification) of routines that enable businesses to benefit from System z’s right-sizing of IT resources to business-critical workloads.
Reality: the staggering complexity of our planet in 2009 probably means “simplification” is both a goal and relative term. System z, while great, requires learning BEFORE its incredible simplification potential gets realized by the user community. But, as we said in the beginning of this edition: “Go slow to go fast.” Think 9,445 transactions per second without a miss. Pretty simple, when viewed that way.
For more information, visit the System z home page.
Yarns about the new realities of the mainframe, spun without spin.
Here is a list of all the entries in this blog/column. You can also scroll down to see all of these, as well.