Storagezilla

Lightning strikes!

Guess what?

Data doesn't just exist on the storage array. And neither does EMC.

The 6th of February.

Here.

The Channel, unleashed.

Joe said it on the earnings call, the Direct Sales Force & the Channel are EMC's primary routes to market.

To prove how important the Channel is and to kick Dell square in the nuts, (Though I'm sure that's entirely coincidental and an unintended side effect) in the UK & Ireland EMC's Direct Sales Force is transferring an addressable market of $3B+ in more than 850 customer accounts straight to Channel partners.

Possibly the largest value unlock in the UK&I IT space for a long long time. 

EMC plans on keeping 150 customers as direct accounts for now but that number will fall over time as some opt to go with the Channel.

The shot at a large sack of money isn't a bad way to start the year if you're an EMC Channel partner.

Reflections in a photographic world

This post is more about business in general rather than storage in specific but it’s worth the discussion.

To use some very primitive descriptors to push an idea, before the advent of photography life paintings captured a period or an age. With the invention of photography we began capturing moments.

As someone who dabbles in photography I did get a slight twinge earlier this week when Kodak filed for Chapter 11 but found it timely as we had more than a few ‘Only the paranoid survive’ conversations of our own inside EMC during our annual sales kick off earlier in the week.

Kodak, which brought photography away from glass plates with George Eastman’s creation of gelatin roll film and then carved out entirely new markets with film canister packaging, was long a benevolent patron of photographers. They did the research in chemistry, optics and electronics which the technology megacorps at the time saw no value in. Kodak didn’t only create the first digital camera they also created Autofocus and point and shoot.

Think about that.

Dean Peterson over at Kodak went to bed one night, got up the next morning and created Autofocus.

That’s a very compressed timeline I know but it didn’t exist until he created it. While you might curse autofocus when it locks on to the wrong point(s) if you’ve ever had to manually focus a camera to grab a shot at a moments notice you know that autofocus tends to tilt the odds of getting the shot in your favour.

For people of a certain age Peterson also created some iconic products for Fisher Price. There was a Fisher Price Audio Tape Recorder in our house in the 80’s and the thing could survive being dropped down stairs and having drinks spilled on it. Two things which frequently happened.

In bringing any product to market you can’t just have a good idea, you need the ability as a company to execute. Even if your idea stinks up the place if you’re an execution machine you can limit the damage and carry on with the next idea. The sign that Kodak’s execution ability had declined during the fat of the land film years came with the launch of APS, the Advanced Photographic Standard. Kodak looking at large boxy expensive digital cameras with their awful resolution and worse optics decided to take on these so called smarter cameras (They weren’t) with smarter film. (It was)

APS not only extended their existing film delivery and processing model, the multi-decade company cash cow, but also added functionality like the ability to embed metadata on the film negative itself. Stuff like date and time the film was shot at, what the aperture and shutter settings were for a shot and so on. It was clever, but there was something far cleverer already on the market. So clever I believe it was the only thing in the century since Eastman started shipping film that was market transforming to photography.

Photoshop.

With the personal computer market booming as Microsoft and Intel looked to put a computer on every desk and in every home, the web exploding with the rise of Netscape, Hotmail giving anyone an email address & attachment quota and early adopters with access to image editing software like Photoshop, getting photos off a camera and into the computer was what people wanted. Kodak backed the wrong horse to the hilt with APS, they made film smarter when they should have been looking at what was going to kill film entirely and they didn’t have the execution ability to make a save when it all went horribly wrong.

The 35mm film market continues today. APS clearly being an answer to a problem photographers didn’t have since 35mm film was perfected over more than 100 years and was proven to just work.  Today APS survives, in a fashion. Those APS-C sized sensors in entry level DSLRs and compact system cameras are named such due to providing a relatively close crop factor to an APS film negative exposed in Classic mode. (The other two APS film modes being High Def and Panoramic) But APS film was a catastrophic failure and film was the marrow in Kodak’s bones.

I do expect Kodak will also survive and exit Chapter 11. It’ll restructure. It’ll shed parts. It’ll sell it’s patent portfolio to pay down debt and probably end up taking a license from the new owner to use the technologies it created, but it’ll be back in one form or another like Polaroid. Polaroid have Lady Gaga designing products for them now, I still have a Polaroid Land Camera and various boxes of expired film around the place so I’ll stick with that for a while.

The lesson I’ve learned from both EMC Kickoff and Kodak are that good ideas are worthless without brilliant execution and if you’re not up to your elbows in what’s going to kill your business then you’re not doing your job right.

Backup Recovery Systems gets even bigger.

I have no idea what EMC did in Q4 or what the year looks like, EMC announces before the bell on the 24th, but it’s clear to me from speaking to people in the field there has been market share expansion across the board. What that looks like and to what extent it has occurred we’ll know next Tuesday.

On the topic of market expansion, today IDC announced that EMC Backup Recovery Systems in the first half of 2011 had 62% of the purpose built backup appliance market. That market growing by 65% year over year even with the economic downturn and contracting budgets.

Since BRS is my division (I say that like I own the bricks in the buildings and the blood in employees) I’ve been lucky to bring a portfolio as intellectually interesting as Avamar, Atmos, Centera, Data Domain, DLm, NetWorker and SourceOne to customers.

All in, by revenue I estimate BRS is the largest backup, recovery and archiving focused player in the business. And while we’re 3x larger than the nearest competitor in the purpose built backup appliances market this year Pat Gelsinger has tasked us with doing even more.

With that in mind, and paperwork still to be processed, I’m hanging up my sales spurs and going to work for BRS CTO Stephen Manley. He has big ideas and the love of the BRS software development teams to make them happen.

After selling the future of backup, recovery and archiving, it’s time to go help create the future of backup, recovery and archiving.

M&A, the EMC way.

As the M&A budget for 2011 appears to have been handed back with no significant purchases, and I thought there would be, it might be worth taking a wide angle look at EMC’s M&A model.

The bottom line is that EMC buys technology and people. It doesn’t buy revenue.

When I started with EMC the company was doing about $5B in business. By the close of this year the company should be doing $19B in business. What’s interesting about those two numbers is that there’s never been a significant boost in revenue coming in with an acquisition.

What do I mean by that? I mean all revenue gains have been as a result of organic scaling alone. If you examine EMC’s acquisition model, when not buying assets at fire sale prices from a failed start ups (Too numerous to list, looks like that scene in Raiders Of The Lost Ark where they’re crating up the Ark and putting it in the massive warehouse of curiosities and dangerous items) or key technology & people in a bidding war with competitors (RSA, Data Domain) the EMC acquisition formula at it’s most basic appears to be a valuation of between 10x to 12x revenues.

VMware was doing about $60M in business when EMC bought it for $635M, Isilon about $200M when it was acquired for $2.25B, Greenplum etc, etc so on and so forth. The common thread is there hasn’t been a single deal to date which brought any meaningful revenue boost in the door on acquisition day. Growth has come as a result of running acquired business well and organically scaling the reach of their technology using EMC’s sales force as a multiplier to create new revenue after the acquisition.

Yes like every other acquisition model you do get some bow wows every now and then, fortunately none of them are ever large enough to put a dent in the universe and you tend to pick up some good people who might find a more fulfilling career somewhere else inside the company. But where this model is in complete opposition to other large scale acquisition models, which tend to buy faltering companies with widely distributed technology but deliver a much larger day one revenue boost, is that you never start on the first day maintaining a massive legacy customer base who’s growth days are over.

So, when EMC looks at moving into a segment of an existing market, or into a market adjacency via acquisition, EMC tends to look at buying the leading technology with good people in a space with large growth opportunities.

Revenue expansion in the short term isn’t the primary concern. The growth opportunity over the longer term is as if you manage that well it’ll deliver meaningful additive revenue for many years to come.         

The future, with Storagezilla.

Not that much of the future but what can I promise you from EMC in 2012?

More.

A lot more in some cases, a bit more in others. I’ve seen the product pipeline, out until the start of 2014 for some and while situations in specific and the market at large changes, what won’t change is the fact that there is more coming.

Next year there will be something for everyone in the audience.

And next year isn’t that far away.

Now. Lets you and I speculate for a moment. I’m of the opinion that NAND Flash is going to be with us well into 2017/2018 before it’s superseded by another solid state technology.

Why that long? Because it takes that long to build an entirely new architecture and deliver it in volume to support whatever comes next.

But before what comes next, ignore the hype of pretenders to Flash’s throne. No replacement technology is close to shipping in a meaningful volume. There’s specific ideas out there being floated as being “just a few years away” which I heard were “just a few years away” when I was in college more than a decade ago. We’re still waiting but I’m not complaining. Turning sand into storage takes a while to get right. 

For the rest of this decade you’re going to use NAND Flash everywhere, and everyone will deliver it packaged as everything over the next few years if they’re not doing it already today.

Used as server cache, used as array cache, used as server storage, used as array storage, used for metadata acceleration, used for whatever. It’s just going to be used and used in volume, packaging be damned.

Assuming a Flash successor does make itself known it’ll be the end of the decade before the operating system vendors and the system builders will embrace it as part of a new technology direction. Which might not be entirely compatible with how we’ve been doing things to date. A lot of the technologies being floated are not interface compatible with current technology.  

But nothing is going to happen until it’s clear to everyone what this superseding technology will be.

So until that moment, where it’s apparent to everyone that a specific technology is the next big thing, NAND Flash will remain both the present and the future of storage media layer technology for the next decade.

PCM Vs Flash

Interesting demo of Phase Change Memory Vs Flash. Still well in the development stage but you can see the promise.

Atmos: Object storage is only the start.

I’ll admit to not being that interested in the Atmos Cloud Delivery Platform up until quite recently.

To me all the action was in the policy driven Petabyte scale multi-tenant distributed storage system. Looked at that, seen it advanced beyond release and I then went off to look at something else. But thinking back on it in totality I found myself asking how do you make that what it needs to be?

How do you make it ubiquitous & invisible? Storage, delivered to end users and metered the same way as water, natural gas or electricity?

Well that’s what the Cloud Delivery Platform does for Atmos.

It provides self service access & storage management of a multi-tenant environment which can be metered on things like utilization or bandwidth, with the ability to integrate into billing systems. And on top of that it’s white label, meaning you can get in there modify and re-skin the entire user experience to ensure continuity with any other services you’re providing to your users.

When you consider that to date all of the functionality in the above paragraph tended to be one shot custom development jobs, where you’d back a truck full of money into your object storage vendors driveway and they’d put a team of their people on it for whatever amount of time it took to get it right, this is a significant step forward. And it’s deployed as a VMware vApp.

Indeed there are some  Atmos customers who’ve deployed the entire shooting match as a vApp, switched on storage as a service with self service access, metering, billing & management in an afternoon on top of vSphere and then went home for the night.

Moving away from the universe of tenants and sub-tenants and focusing on Application Developers for a moment, there are monthly developer workshops running where developers can learn about the system and stand up their own applications enabled for Atmos in an EMC Development Cloud.

Developers want code, not trips to the Airport to go sit in a class room. Do it all online with other developers.

What about users?

I signed up with Oxygen Cloud recently and found it pretty useful for Sync & Share.

I’m using EMC’s Atmos deployment and Oxygen Cloud across my PCs, iPad and Android phone, with the collaboration options Oxygen Cloud allows for files enabled, so I can work with a distributed team.

With the amount I read and the feedback I give It’s better than email attachments, trust me.

 

All in, I had Atmos ass backwards for a while now. I thought the most interesting part was the storage system itself, when all the action was really in self service, was in the ease of application enablement and was in what third parties have been building using both of those.

A bigger picture for a bigger distributed storage system.

EMC & F5: Deduplication Storage In A Tiered Storage Environment

Seamlessly Integrating Deduplication Storage into a Tiered Storage Environment

By seamlessly integrating deduplication storage into a tiered storage environment, you can reduce the complexity of data management while driving down capital and operational expenses significantly.

Join F5 and EMC and learn how you can reduce the cost and complexity of managing unstructured data. 

via resources.f5.com

I might give this a look myself to find out what's new with F5 & Data Domain.

Drought.

First it was the point of sale price hikes, then today I was browsing around a major high street retail computer store when I spotted this printed out and scattered amongst the computers and external storage devices.

A visit to two competing retailers across the city confirmed a shortage of external hard drives with only the higher end Media & Home NAS devices on the shelves. When I asked retail assistants if they were due to receive any new stock soon one told me they were informed by their parent company that any deliveries before the end of the year should be considered a stroke of good luck.