Search results matching tags 'BladeSystem' and 'blade infrastructure'

Chargeback Glitch

Daniel Bowers — Fri, 15 May 2009 16:25:00 GMT

Here’s a story about unexpected behavior prompted by some chargeback rules at Pfizer. (By 'chargeback' I mean the budgeting scheme where IT resources are metered, then costs are attributed to the business units that consume them.)

Larry Cannell notes that when department budgets were squeezed, users moved Sharepoint files to non-Sharepoint systems "all to avoid internal chargebacks. In the end...a lot of extra work...but no impact on the bottom line." For HP Bladesystem, tools that let IT departments implement chargeback are nothing new. There are ways to both measure resource utilization, and to implement high-level chargeback policies.

HP has even added some ground-breaking capabilities here -- bringing continuous capacity planning tools into the x86 space, for example. We’ve also added server power consumption as a resource that can be tracked historically for chargeback calculation – or even capped so businesses can predict their spend.

You have to think like a non-IT guy, though, to predict how your internal customers will react to your chargeback rules. Put yourself in the shoes of a line-of-business manager:

People will order from the "Value Menu". If someone thinks he could squeak by with some minimal level of service, he’s going to try.
Square pegs will wind up in round holes. To an HR guy, using a USB key as the primary backup device for your employee payroll database seems reasonable.
Exceptions have legs. That handshake agreement to charge a little less because the legal department promised they’d use minimal bandwidth? Ten minutes later, every other department will demand “the “low bandwidth deal”.
Don’t put nametags on hardware. Especially with bladed infrastructure, you need to use language that avoids the impression that groups have “bought a server”. They’re paying you for a service…How you deliver that is up to you, not them.

Least Favourite Question

Tony Harvey — Wed, 13 May 2009 21:15:00 GMT

I'm sitting here thinking about writing my first Blog post and trying to come up with something to say. So I figured I'd start by trying to answer one of my least favourite questions (and before you all start to correct my spelling I'm not originally from the USA) and explain why it's so hard to answer.

The Question: "So much power does a blade enclosure use?"

The answer: "It depends."

What does it depend on? Everything.

How many Blades and which Blades do you have in the enclosure, which CPUS, are they the 50W, 60W, 80W, 95W or 120W versions, how many DIMMs and what size and rank are they, which mezzanines are installed, which Interconnects are installed, how many fans, what's the ambient temperature, what applications are running and how heavily loaded are they?

Even if you gave me all this information I still couldn't answer with any degree of accuracy, those final two items applications and application load really do have such a huge impact it makes it almost impossible to give the right answer. The best that I could do would be to give the maximum and minimum power usage based on that configuration and say you'll be somewhere in-between those two values.

In the next few posts I'll go into some detail about this starting with the affect hardware configuration has on power consumption.

Designing infrastructure the way IT wants to work

Gary Thome — Wed, 22 Apr 2009 07:46:00 GMT

We've been on the Adaptive Infrastructure journey at HP for several years now. This week we are announcing an important milestone: BladeSystem Matrix. We've been really thinking a lot about how customers use IT and ways we can optimize IT infrastructure to make it work better for them. We recognize that infrastructure exists for applications, which exist for the business. So we've taken a business and application perspective on how an infrastructure ought to operate.

Deploying an application typically requires an IT architect or team of architects to carefully design the entire infrastructure - servers, storage, network, virtual machines - and then hand off the design to a team of people to deploy, which typically takes several weeks. This length of time is mostly an artifact of the way IT infrastructure is designed. So we decided to change this with BladeSystem Matrix. Now an architectural design is saved out as a template - servers, storage, virtual machines, network, server software image. Then when it is time to provision an application, it's as easy as saying "make it so" - and in a matter of minutes, the Matrix's converged virtualized infrastructure is automatically configured and the application is ready to run. In other words, the way it ought to be.

BladeSystem Matrix is the culmination of several years work at HP - creating an Adaptive Infrastructure that is simpler to buy, deploy and keep running optimally. Applications are easier to provision, maintain, and migrate. We've spent years proving out this architecture, not just in our labs but in real-world environments, with BladeSystem, Virtual Connect, and Insight Software - so we could learn how IT really operates - and more importantly - how it ought to operate.

Some people tell me Matrix's virtualization sounds sort of like a mainframe. Others say that the portal interface reminds them of cloud IT. I guess in a way they are all correct. But unlike those environments, Matrix will run off-the-shelf x86 applications. So I guess I've decided that Matrix is it's own thing.

Did we miss something?

Gary Thome — Mon, 16 Mar 2009 21:35:00 GMT

Every time a competitor introduces a new product, we can't help but notice they suddenly get very interested in what HP is blogging during the weeks prior to their announcement. Then when the competitor announces, the story is very self-congratulatory "we've figured out what the problem is with existing server and blade architectures". The implication being that blades volume adoption is somehow being constrained by the very thing they have and everyone else is really stupid.

HP BladeSystem growth has hardly been constrained; with quarterly growth rates of 60% or 80% and over a million BladeSystem servers sold. So I have to wonder if maybe we already have figured out what many customers want - save time, power, and money in an integrated infrastructure that is easy to use, simple to implement changes, and can run nearly any workload.

Someone asked me today "will your strategy change?" I guess given the success we've had, we'll keep focusing on the big problems of customers - time, cost, change and energy. It sounds boring, it doesn't get a lot of buzz and twitter traffic, but it's why customers are moving to blade architectures.

Our platform was built and proven in a step-by-step approach: BladeSystem c-Class, Thermal Logic, Virtual Connect, Insight Dynamics, etc. Rather than proclaim at each step that we've solved all the industry's problems or have sparked a social movement in computing; we'll continue to focus on doing our job to provide solutions that simply work for customers and tackle their biggest business and data center issues.

9 Trends in 2009: What's Hot and What's Not in the Data Center

newtonja — Thu, 22 Jan 2009 13:30:00 GMT

We can’t let Illuminata, Gartner, IDC and Forrester have all the fun, so we sat down and came up with our own "What’s Hot and What’s Not in the Data Center" list for 2009.

NOTE: This blog post will self-destruct on December 31, 2009 should anyone feel the need to analyze our prognosticating skills in 2010.

1. Power as a resource is HOT. Power as a commodity is NOT. If you knew your old refrigerator in the garage was sucking fifty bucks in juice a month, you’d pitch it or replace it, right? The problem is; you have no idea how much power it costs you. In 2010, you’ll never think about power in the same way. It’s no longer just a spigot of electrons with a bill that goes to the suits upstairs. Power is a precious resource to your data center and a big part of your budget that stands in the way of growth in 2010. “You can’t manage what you can’t measure”, so 2009 is the time to start measuring your power usage in detail so you understand what you need, what you have and what you’re wasting.

2. TCO is HOT. TCO is NOT. Huh? TCO will be reprioritized in 2009. Take Cost Out is the new TCO. Okay, we don’t want to overplay this one. Of course you want to be as efficient as possible down the road once the 2009 storm passes. However, if you’re ever going to get there, you have to take cost out now. That’s going to mean you have to make tough choices and some big leaps forward in order to put in place an infrastructure that can deliver savings today and be ready for tomorrow. We believe very few stones will be left unturned in 2009 as businesses scourer their data center to find hidden pockets of cost – cables, steps in processes, HA, fibre channel, aging servers, wasted watts, unused A/C – nothing can hide from the new TCO. Those trying to limp through 2009 by patching up some aging technologies will find themselves in world of hurt in 2010.

3. Knowing is HOT. Guessing is NOT. Whether you’re talking capacity planning for your apps and virtual machines, the power and breaker size you need per rack or the storage for your data explosion, using the old ‘rules of thumb’ for quarterly budgets aren't going to cut it in 2009.  Getting better data out of every circuit board that you can then use to take informed action will be critical to justify growth and to help you squeeze the most cost out from your consolidation projects.

4. Packaged infrastructure is HOT. Piecemeal infrastructure is NOT. Sorry IBM, the mainframe isn’t part of this one. We’re talking about pooled and shared infrastructure based on industry standard components. We think you’ll see more packaged infrastructure solutions tailor-made to different applications and environments whether it’s a rack at a time for mega clusters or a unified communication platform for a small branch office. You already see it with ExDS, BladeSystems, PODs, NeoView, NonStop blades; the trend is probably already here but it’s going to really take off in 2009. The idea is simplified delivery, integration and expansion. You just won’t have the time in 2009 to try and figure out how to get widget A to talk to widget B.

5. Unified is HOT. Siloed is NOT. Whether it’s Cisco, Microsoft, IBM or us, the vision of unified infrastructure is clearly our shared goal. We just have different names for it. The only question is how do we make progress in 2009? We know one thing for sure; you can’t get there by forcing the perspective one silo one on another. Network packets won’t unify your infrastructure any more than processor architectures will. The only path to the unification you seek is from the top down starting at business and application services. Understanding and managing in a unified way provides a different perspective on what tomorrow’s infrastructure looks like. When you recognize that, you see that the center of the universe can’t be found inside the network, the storage or the servers. It’s at the business level.

6. Performance per sq ft, per dollar per watt is HOT. Moore’s Law is NOT. The days of chasing the tail of processor performance are quaint, but the new global economic reality will create a whole new class of benchmarks to help you better compare your choices. Whether your issue is space, power or cost, you’ll be empowered in 2009 with a whole new set of much more relevant benchmarks to see where you stand. SPECPower, VMmark and others are just scratching the surface of what will be a renaissance in data center metrics.

7. DAS is HOT. SAN is NOT. Okay, okay. The SAN isn’t going anywhere. But there will be a lot more choices in 2009 that flip the economics of storage on their head and put server admins in more control of their storage needs. Last night I was browsing for some home storage backup and came across a deal for a 1TB home back-up for $149. Buying your first TB in a traditional SAN will set you back $30 to $50k. (Calvin Z's going to kill me). Basically, storage is delivered by drives. Shouldn’t you be able to pile up all the drives you have, DAS or otherwise and carve up that capacity how you see fit? Check out some of the cool stuff we can do now with LeftHand’s innovations and you’ll see what we mean.

8. Virtual infrastructure is HOT. Virtual machines are NOT. Or said another way, “Virtualization is dead! Long live virtualization”. 2009 will shift priorities from optimizing server capacity with virtual machines to looking for new opportunities at the server edge to extend the savings and consolidation to the network, management and storage realms. Virtual infrastructure will be the new mantra and managing it, coordinating it and aligning it to the business will be the key. In 2009 more people will think differently about infrastructure as service and something that you simply carve up and allocate capacity based on your demands. It aligns to you, not the other way around. With this in place, automation starts getting real too!

9. Dynamic Core Utilization is HOT. Multi-core apps running one application is NOT. This one almost fell to runner up status simply because it was a mouthful and a little geeky, but we needed 9 things. Seriously though, the flexibility to adjust core utilization to match a workload has been a long time coming for the x86 world. As 2009 starts to move us beyond quad-core processors, it makes no sense to continue the old, tried-and-true practice of one app per server.

The runners up:

Dynamic Power Capping is HOT. Power Face Plates are NOT. This one can be summarized above with “Knowing vs. Guessing”. The low hanging fruit for reducing power consumption is just about gone. It’s going to take more intelligence and coordination at the rack, row and datacenter level. The ability to reclaim data center capacity simply be allocating only the power you need makes too much sense to not make our list.

Converged Fabrics are HOT. Silo’d I/O traffic is NOT. Talk about an oldie but a goody, this one might start to make it over this hump this year. Aggregation of I/O to a single physical layer ushers in a whole new opportunity to simplify and cut big costs. See most of the bigger trends we mentioned above and you see that this one is key.

Industry standard gear for Telco is HOT. Telco-only gear is NOT. Ah, one of the last bastions of proprietary gear. This one has been predicted so often, it fell to runner-up. But the slow march continues and we think 2009 will speed things up a lot as more telcos start to see the benefits of gear like blades and standard rack servers on their balance sheet.

Battery back-up at the rack is HOT. UPS rooms are NOT. Nice idea that just doesn’t make as much sense as folks thought. Hogs of data center floor space, budget and that nasty little 10% loss in efficiency make this one at least worth of the runner-up list.

Blades are HOT. Mainframes are NOT. Give me a break. We are the HP Blade Team. It just wouldn’t be an IT hot/not list without one little jab at poor Big Blue. ;-)

We’d love to hear from you. Tell us what you think about this list and share what’s on your HOT and NOT to-do list for 2009.

BladeSystem Supercomputers

Gary Thome — Mon, 01 Dec 2008 17:34:00 GMT

The latest Top 500 supercomputer list just got released. BladeSystem c-Class was well represented once again. With 201 entries (40.2%) of the top 500, it has the most entries of any product line. The ever popular ProLiant BL460c made up the most entries, but we also had strong showings of the BL465c and the two-in-one blade the BL2x220c, and the BL685c 4-way blade made a showing as well. BladeSystem supercomputers are used for university and government research, weather modeling, semiconductor development, automotive, telecom, IT services, web infrastructure, financial services, rendering, and many other applications. I'm sure Dell was excited with their new blades line as well. Since they like to compare their blades with HP BladeSystem, I thought I would share how the two compared on the Top500 list:

HP BladeSystem had 199 more entries in the top 500 list than Dell's new blades
HP had 39.8% more share than Dell's new blades (40.2% vs. 0.4%)
HP had 100.5 times more entries than Dell's new blades

Dell's new blades accounted for two entries. Congratulations Dell!

Okay enough of these comparisons. We're excited to see so many customers from Audi to Zeta and lots of customers in between using BladeSystem. If you would like to see a listing of companies building supercomputers with BladeSystem, go check out the top 500 website listing and sort by vendor.

10Gb Ethernet: Divide and Conquer with Virtual Connect Flex-10

Gary Thome — Mon, 17 Nov 2008 23:53:00 GMT

Moore's law has usually been used to predict general trends in semiconductors. While not exactly a perfect analogy, we have seen trends in interconnect bandwidth increase. Ethernet has seen bandwidth increase ten-fold every few years, with the latest transition to 10Gb. Usually these transitions take a while because the costs to transition are high, and the transition to 10Gb has followed that trajectory - until now.

Yesterday we announced an exciting new technology: Virtual Connect Flex-10. We've figured out a way to deliver 10Gb Ethernet technology at a price lower than what many people are spending on 1Gb technology today. As a result we can help customers get onto 10Gb technology sooner than they otherwise could have done before.

We've noticed that many customers are buying four or more NICs for their servers, sometimes due to bandwidth constraints, other times due to network segmentation or security constraints, and usually for redundancy. As a result, customers spend an awful lot on a bunch of 1Gb networks. We figured out that we could help customers by providing a 10Gb network connection that can be divided into up to four connections, replacing the need for up to four NICs. By doing this we conquer the high price for 10Gb Ethernet connectivity by delivering up to 8 network connections at costs that are less than what many customers pay for four 1Gb connections. At the same time they can allocate more bandwidth for one or more links or maintain the multiple connections they want for security reasons, or do a combination of both. And to top it all off, Virtual Connect takes less space and power too. We think this is very cool.

Bottom line: more bandwidth, more flexibility, less costs and less power. More of what you want and less of what you don't want. We think this is a good combination. This is why we believe the quickest, most affordable way to move to 10Gb is Virtual Connect Flex-10, and the time to do it is now.

When is choice not choice?

Gary Thome — Thu, 11 Sep 2008 14:36:00 GMT

Following on the heels of our virtualization launch last week, Dell made a virtualization announcement of their own yesterday. They announced a variety of third party products they now support and re-announced the blades servers they introduced last week, but this time referring to their virtualization design. Curiously they compared their two-socket M805 full-height 16-DIMM blade to our four-socket blades, ignoring our two-socket half-height 16-DIMM ProLiant BL495c virtualization blade announced last week. I guess comparing their blade against an HP blade that is half the size wouldn't have sounded as good.

But what really caught my attention was their statement in their press release that their strategy is "grounded in choice". I imagined how this strategy plays out with blades:

Customer "I'd like to choose a UNIX blade please."
Dell does not offer this choice.

Customer: "I'd like to choose a storage blade please."
Dell does not offer this choice.

Customer: "I'd like to choose a workstation blade please."
Dell does not offer this choice.

Customer: "I'd like to choose a half-height blade with 16 DIMM sockets please."
Dell does not offer this choice.

Customer: "I'd like to choose a two-servers-in one blade for my grid app please"
Dell does not offer this choice.

Customer: "I'd like to choose a Non-stop blade" please.
Okay I could go on, but you get the picture.

As it turns out a lot of customers want these kids of choices. Why? Because a blade everything strategy means they can get the time, energy and cost savings BladeSystem offers for more of their IT infrastructure. They can have a simpler, more consistent way to deploy, maintain, manage and service their infrastructure. But here again Dell has clearly differentiated themselves, stating that "We are not blade everything". I guess this is one choice Dell does not want to offer to customers.

Games vendors play . . . with power efficiency claims

Gary Thome — Mon, 18 Aug 2008 21:13:00 GMT

These days, server power efficiency is top of mind for everyone. Five years ago, most customers had no idea how much power a server used. Now, everyone knows - it's a lot. In some cases, customers are making power consumption the primary criteria for vendor selection.

So why is it so dang hard to get a straight story on exactly what options are the most power efficient?

In our experiences with power measurements, we found lots of ways to the results get skewed; whether intentionally or accidentally. We honestly try to avoid them, but here are some of the dirty little secrets a lot of vendors don't want you to know about their power testing results. Consider these red flags next time someone spends lots of money in the Wall Street Journal to post a big claim with lots of fine print on power savings.

Lab Queens

One easy way for vendors to skew power results is to cherry-pick low power components. Processors (even those in the same power grade) and memory DIMMs tend to consume significant power and can have wide variances in power consumed from one part to another.

We call units built by cherry-picking components "Lab Queens".

Generally, these do not represent what a customer might actually be able to purchase. Here's an example where we tried to avoid this scenario; the systems compared on this competitive power report on our website uses the exact same processors and DIMMs on all units tested, thereby eliminating differences due to component variances. It would have been easy to create a Lab Queen and bump the results higher, artificially.

Configuration Errors

Peak power efficiency for a given system always requires the "best" configuration. Memory power is mostly a function of DIMM count, and not capacity. Therefore, to achieve minimum power, the minimum number of DIMMs should be used to realize the desired memory capacity.

If you see a comparison where the DIMM count is different or not mentioned, you might notice the strong odor of dead fish.

On the infrastructure side of the equation, a blade system can be configured with varying numbers of fans and power supplies. Obviously some configurations will produce better power efficiency than others. For instance, 4 power supplies run more efficiently than 6, assuming that 2+2 power redundancy delivers adequate power for a given configuration. The most power efficient configurations though won't always be generated by a competitor when publishing a comparison. Honestly, there are such big differences here that an apple to apple comparison is tough. Only an HP BladeSystem dynamically throttles fan speeds based on demand across the different cooling zones in an enclosure.

Environmental Differences

The voltage level and room temperature can cause variances in the power supply efficiencies and the speed the system fans must run. Consequently it is important that when comparing servers to verify that these parameters are held constant. I once heard a story of an engineer setting up a box fan to blow air on his servers to try to reduce the fan power of the blades.

Obviously he wasn't planning on counting the power from the box fan!

Remarkably, there are published benchmarks that allow for external cooling to be deployed and not counted as a part of the system power.

Gaming the Benchmark

Any benchmark a vendor publishes is necessarily narrow in what it measures. When looking at a published benchmark, you should consider how close the benchmark mimics your applications. If the benchmark is similar to your applications, the results might be relevant to you. If not, then you might do best just to ignore them. Also, some benchmarks are very loose on configuration requirements, making it very easy to "game them" and making the results all but totally useless. One benchmark that is broadly published in the industry falls into this category.

Where to go from here?

At this point, you are probably thinking "Wow! This is much more complicated than I really thought!" And you are right. Benchmarking fairly is very tricky business, even if you are trying to be fair. The best results you can get are the ones you measure yourself. In my next blog post, I'll comment on some of the techniques and challenges with measuring power yourself.

One million and counting

Gary Thome — Fri, 11 Jul 2008 14:16:00 GMT

1,000,000. One million. 10⁶. No matter how you write it, that is a big number. I've never tried to count to one million. But now we have sold our millionth blade server; which is cause for reflection.

We created our blades products on the conviction that there was a better way to make IT infrastructure work. We knew from customers that deploying, managing and changing IT was challenging, time-consuming, expensive, and power hungry. It was this desire to help solve these challenges that lead us to create the BladeSystem that we offer today.

People's opinions vary on why customers have bought over a million blades from HP. Some people think it stems from Shorty's presidential run (after all there is some attraction to a candidate with a proven track record of saving time, cost, and energy!). Others think that BladeSystem is just a cool product. Perhaps it is cutting-edge products like our 2-in-1 ProLiant BL2x220c blade server.

As for me, I think customers like the time-saving, cost-saving, and energy-saving technologies built into BladeSystem. They like having a single point of infrastructure from which they can deploy their various servers and applications. They like the intuitive way BladeSystem is deployed and managed.

In any case, we are glad that we are able to help customers. We are happy whenever BladeSystem can be a part of addressing their biggest pain points. We are excited to be the first to achieve this milestone. But we are not done yet on helping customers reduce costs and energy; to make BladeSystem faster to deploy, manage and change.

So hang onto your hats - we have a lot still to come!

Gary