In his 1956 dystopian sci-fi novel “The City and the Stars”, Arthur C. Clarke puts forth the fundamental design tenet for making eternal machines, “A machine shall have no moving parts”. To someone from the 1950s current computers would appear to come close to that ideal – the CPUs and memory perform silent magic and can, with some ingenuity, be passively cooled, and invisible electronic signals carry information in and out of them to networks and … oops, to rotating disks, still with us after more than five decades[i]. But, as we all know, salvation has appeared on the horizon in the form of solid-state storage, so called flash storage (actually an idea of several decades standing as well, just not affordable until recently).
The initial substitution of flash for conventional storage yields immediate gratification in the form of lower power, maybe lower cost if used effectively, and higher performance, but the ripple effect benefits of flash can be even more pervasive. However, the implementation of the major architectural changes engendered across the whole IT stack by the use of flash is a difficult conceptual challenge for users and largely addressed only piecemeal by most vendors. Enter IBM and its Flashahead initiative.
What is Happening?
On Friday, April 11, IBM announced a major initiative, to the tune of a spending commitment of $1B, to accelerate the use of flash technology by means of three major programs:
· Fundamental flash R&D
· New storage products built on flash-only memory technology
For the vast majority of Forrester customers who I have not had the pleasure of meeting, my name is Henry Baltazar and I'm the new analyst covering Storage for the I&O team. I've covered the Storage industry for over 15 years and spent the first 9 years of my career as a Technical Analyst at eWEEK/PCWeek Labs, where I was responsible for benchmarking storage systems, servers and Network Operating Systems.
During my lab days, I tested hundreds of different products and was fortunate to witness the development and maturation of a number of key innovations such as data deduplication, WAN optimization and scale-out storage. In the technology space "Better, Faster, Cheaper - Pick Two" used to be the design goal for many innovators, and I've seen many technologies struggle to attain two, let alone three of these goals, especially in the first few product iterations. For example, while iSCSI was able to challenge Fibre Channel on the basis of being cheaper - despite being around for over a decade many storage professionals are still not convinced that iSCSI is faster or better.
Looking at storage technologies today, relative to processors and networking, storage has not held up its end of the bargain. Storage needs to improve in all three vectors to either push innovation forward, or avoid being viewed as a bottleneck in the infrastructure. At Forrester I will be looking at a number of areas of innovation which should drive enterprise storage capabilities to new heights including:
We all know the conventional wisdom about cloud computing: it's cheap, fast and easy. But is it really that much cheaper? Or is it simply optics that make it appear cheaper?
Optics can absolutely change your perception of the cost of something. Just think about your morning jolt of coffee. $3.50 for a no-foam, half-caf, sugar-free vanilla latte doesn't seem that expensive. It's a small daily expense when viewed by the drink. It appears even cheaper if you pay for it with a loyalty card where you don't even have to fork over the dough and the vanilla shot is free. But what if you bought coffee like IT buys technology? You would pay for it on an annual basis. That $3.50 latte would now be about $900/year. For coffee? How many of you would go for that deal? That's optics and it plays right into the marketing hands of the public cloud services your business is consuming today.
But optics aside, is that $99/month per user SaaS application just another $20,000 per year enterprise application? Is that $0.25 per hour virtual machine just another $85 per year hosted VM? No, it's not the same. Because the pricing models are not just optics but an indication of the buying pattern that is possible. If you buy it the same way you do traditional IT, then yes, the math says, there's little difference here. The key to cloud economics is to not buy the cloud service the same way you do traditional IT. The key to taking advantage is to not statically and rotely consume the cloud. Instead, consume only what you need when you need it — and be diligent about turning off when you aren't.
This week, the New York Times ran a series of articles about data center power use (and abuse) “Power, Pollution and the Internet” (http://nyti.ms/Ojd9BV) and “Data Barns in a Farm Town, Gobbling Power and Flexing Muscle” (http://nyti.ms/RQDb0a). Among the claims made in the articles were that data centers were “only using 6 to 12 % of the energy powering their servers to deliver useful computation. Like a lot of media broadsides, the reality is more complex than the dramatic claims made in these articles. Technically they are correct in claiming that of the electricity going to a server, only a very small fraction is used to perform useful work, but this dramatic claim is not a fair representation of the overall efficiency picture. The Times analysis fails to take into consideration that not all of the power in the data center goes to servers, so the claim of 6% efficiency of the servers is not representative of the real operational efficiency of the complete data center.
On the other hand, while I think the Times chooses drama over even-keeled reporting, the actual picture for even a well-run data center is not as good as its proponents would claim. Consider:
A new data center with a PUE of 1.2 (very efficient), with 83% of the power going to IT workloads.
Then assume that 60% of the remaining power goes to servers (storage and network get the rest), for a net of almost 50% of the power going into servers. If the servers are running at an average utilization of 10%, then only 10% of 50%, or 5% of the power is actually going to real IT processing. Of course, the real "IT number" is the server + plus storage + network, so depending on how you account for them, the IT usage could be as high as 38% (.83*.4 + .05).
Our latest survey on IT budgets and priorities shows that 35 percent of enterprises have a big focus on cloud computing (calling it a high or critical priority), but do we really know how best to apply that investment?
We continue to see a large disconnect between what the business wants from cloud services and what the IT department wants to offer and support. The short story is the business wants public cloud services (or something very, very close to this value proposition) for delivering new services and capabilities to market. Yet IT wants to offer private cloud solutions that improve operational efficiency and drive down overall IT costs. IT doesn't have its head in the sand about business' demands, they just have to balance these desires against what IT is measured on - the cost and security of services provided. And frankly they don't trust the public cloud.
Knowing the psychology above, how best can an enterprise set a winning cloud strategy? if it invests purely against the business care-abouts it may win time to market but risks investing ahead of its ability to support and protect the business. If it invests against the IT priorities it risks alienating the business, increasing circumvention and being a laggard competitively. The answer lies in striking an appropriate balance between these conflicting priorities and choosing a strategy that encourages the most collaboration between business and IT and accelerating everyone's experience level with these new technologies. And that balance will be different for every firm based on their competitive market, regulatory environment and geography. But in general, most enterprises are being far more conservative than they should.
As developers, we often ask for more resources from the infrastructure & operations (I&O) teams than we really need so we don't have to go back later and ask for more - too painful and time consuming. We also often don't know how many resources our code might need, so we might as well take as much as we can get. But do we ever give it back when we learn it is more than we need?
On the other hand, I&O often isn't any better. The first rule we learned about capacity planning was that it's more expensive to underestimate resource needs and be wrong than to overestimate, and we always seem to consume more resources eventually.
Yesterday, Amazon launched an adjunct to its successful Amazon Web Service (AWS) elastic cloud offering. While we don’t normally comment on every product release, this one is significant — primarily because of who is doing it. The Simple Workflow service (SWF) clearly has nothing to do with Adobe’s Flash offering (although techno-nerds may initially think so, given the acronym).
So what was this all about? The business model is certainly interesting: an elastic, configurable workflow capability that’s distributed across any number of access points. Essentially, this will allow an organization to orchestrate processes in the cloud, linking participants up and down the value chain.
“Amazon Simple Workflow Service (Amazon SWF) is a workflow service for building scalable, resilient applications. Whether automating business processes for finance or insurance applications, building sophisticated data analytics applications, or managing cloud infrastructure services, Amazon SWF reliably coordinates all of the processing steps within an application.”
Pricing is initially free and then transitions into a blended, low-cost consumption model, with charges oriented around execution steps, bandwidth usage, how long the task is active, and signals/markers, etc. With usage charges at around $0.0001 per execution step, this gives you an idea of how small the operating overhead might be.
Cloud – people can’t agree on exactly what it is, but everyone can agree that they want some piece of it. I have not talked to a single client who isn’t doing something proactively to pursue cloud in some form or fashion. This cloud-obsession was really evident in our 2011 technology tweet jam as well, which is why this year’s business technology and technology trends reports cover cloud extensively. Our research further supports this – for example, 29% of infrastructure and operations executives surveyed stated that building a private cloud was a critical priority for 2011, while 28% plan to use public offerings, and these numbers are rising every year.
So what should EAs think about cloud? My suggestion is that you think about how your current IT strategy supports taking advantage of what cloud is offering (and what it’s not). Here are our cloud-related technology trends along with some food for thought:
The next phase of IT industrialization begins. This trend points out how unprepared our current IT delivery model is for the coming pace of technology change, which is why cloud is appealing. It offers potentially faster ways to acquire technology services. Ask yourself – is my firm’s current IT model and strategy good enough to meet technology demands of the future?
Over the past months server vendors have been announcing benchmark results for systems incorporating Intel’s high-end x86 CPU, the E7, with HP trumping all existing benchmarks with their recently announced numbers (although, as noted in x86 Servers Hit The High Notes, the results are clustered within a few percent each other). HP recently announced new performance numbers for their ProLiant DL980, their high-end 8-socket x86 server using the newest Intel E7 processors. With up to 10 cores, these new processors can bring up to 80 cores to bear on large problems such as database, ERP and other enterprise applications.
The performance results on the SAP SD 2-Tier benchmark, for example, at 25160 SD users, show a performance improvement of 35% over the previous high-water mark of 18635. The results seem to scale almost exactly with the product of core count x clock speed, indicating that both the system hardware and the supporting OS, in this case Windows Server 2008, are not at their scalability limits. This gives us confidence that subsequent spins of the CPU will in turn yield further performance increases before hitting system of OS limitations. Results from other benchmarks show similar patterns as well.
Key takeaways for I&O professionals include:
Expect to see at least 25% to 35% throughput improvements in many workloads with systems based on the latest the high-performance PCUs from Intel. In situations where data center space and cooling resources are constrained this can be a significant boost for a same-footprint upgrade of a high-end system.
For Unix to Linux migrations, target platform scalability continues become less of an issue.
During the recent AT&T Analyst Conference, we analysts were all treated to 2 great shows - a tour of Yankee Stadium’s ICT systems (followed by an opportunity to watch the Yankees beat the Royals) and Kevin Peter’s annual spectacle of spherical geodesic references to make sure we all got one thing straight - AT&T’s network delivers value.
Following last year’s event, I referred to the clear presence of AT&T Labs innovations (http://blogs.forrester.com/henry_dewing/10-05-24-att_business_solutions_drives_technological_innovation_practical_use) delivering value to customers. Keith Cambron, President and CEO of AT&T Labs, focused more this year on the importance, value, and urgency of the IPV6 transition, before Kevin delivered another rousing speech declaring “IT’S THE NETWORK!” more frequently and more emphatically as he went. While Kevin’s presentation was all style, he is a serious technology executive having received a master's degree in Information and Technology Management from the Stevens Institute of Technology. Kevin’s stump speech was followed by technical topics and serious presentation styles that illuminated his points.
John Donovan, AT&T’s CTO, settled the audience down describing AT&T plans to deliver solutions using
AT&T’s Network as a ubiquitous, reliable, intelligent platform.