thoughts on computing

A while back there was a flurry of activity around a startup proposing floating datacenters – at the time I thought it was kind of a dumb idea … followed by another post with a more on-the-point headline:

“Floating Data Centers Miss the Point, Add a Bunch of Risk, and Will Keep You Up At Night; On the Other Hand, Deploying Your Applications on a Cloud of Commodity Computers With Appistry’s Application Fabric Will Deliver the Goods” (note: slight edit)

Since then the idea has been rather quiet, though today Engadget is reporting that Google has filed a patent app covering their take on the same notion.

At first glance this appears to be a more far reaching version of the floating data center notion, adding an interesting (though still fairly conceptual) energy generation notion.

The idea of self-generating power ups the potential benefits idea (beyond cooling and portability) a potential big step … yet the two biggest hurdles remain.

Hurdles That Remain
Leaving expense aside, how to connect sufficient bandwidth to a floating data center is remains an enormous challenge – whether wireless or wired, it’s just going to be difficult.

Second, this really will need to be at least as sturdy as The Unsinkable Molly Brown if it’s going to have any value beyond conceptual bantering.

Whether from terrorists, storms, or just inexplicable mistakes, the prospects of all those computers ending up wet wet wet is a sobering one, indeed.

Just ask Mazda … and their problem was only dented cars.

A Big Idea
Bottom line, I think this is a “big idea” that will pop back up from time to time, and will probably even have some very flashy demos and prototypes.

But … and this is a very big but … I think it’s days as a practical alternative for hosting stuff that we really care about are still a long way off … if ever.

Update 1: This story also showed up on Slashdot, with a decent discussion following.

Tonight’s St. Louis pitch event is shaping up to be a lot of fun, and has now grown to include three steps. Even if you haven’t signed up, come on down and join in the fun.

Pitch. If you’re willing to stand up and give a very short pitch (max 3 minutes), show up by 4.

Watch. If you want to watch the best of the pitches, but don’t have your own, show up at 5.

Hang Out. If you just want to hang out and meet other folks who just love starting stuff, then we’re going to be at Schalfly’s Tap Room at about 7:30 or so.

We’re open to tweaking the format of course … the main point is to bring together the startup community in and around the St. Louis area together. There’s a lot of good stuff going on around the area already, and I’m excited about the prospects for more to come.

For more information go here or check out Puredanger’s post as well.

See you tonight!

Nova Spivack, founder and CEO of Twine has an interesting post at RWW. He makes a number of points

The desktop of the future is going to be a hosted web service

The Browser is Going to Swallow Up the Desktop

The focus of the desktop will shift from information to attention

Users are going to shift from acting as librarians to acting as daytraders.

The Webtop will be more social and will leverage and integrate collective intelligence

The desktop of the future is going to have powerful semantic search and social search capabilities built-in

Interactive shared spaces will replace folders

Sorting through a couple of Capt. Obvious points (“The Webtop will be more social” … you’re kidding me?!?!?), he does make a few more interesting points. For example,

The focus of the desktop will shift from information to attention

is a really good point. Sure it’s just the latest way to say “drinking from a firehose”, yet it at least cleanly articulates what we all deal with daily, at levels that when we step back and think about it are nearly incomprehensible … with much more to come.

A Bit of Wishful Thinking
Yet from my perspective at least a couple of the points just fall into true blue believer wishful thinking … as in “But it JUST HAS to happen this way … doesn’t it?”

Ummmm … no.

Let’s pick one to illustrate:

The Browser is Going to Swallow Up the Desktop

That meme has been going around for quite a while. Probably the most famous, recent, and all around hard to escape incarnation of that philosophy is clearly the iphone. So let’s take a look there and see what we can learn.

Great browser? Check.

Uber-outstanding display? Check.

Tons of mindshare with a maximum mind-control field targeted at making everyone believe that browser apps constituted everything anyone would ever need? Check.

Fast network? Check.

Ubiquitous? Check. Check. Check.

On the Road to Web-App Total Domination
Well we all know what happened with six months of this strategy … Google unveiled Android, and Apple had no choice but to open up their platform (ok, not really very open … but at least non-Apple employees can sort of write apps!).

The marketplace is voting at a furious pace, with more than 60,000,000 apps downloaded in the first month. Yes some folks extrapolate from their own first month experience and say that all of this will die down soon, to be replaced by the browser alone.

Yet I just don’t see it.

The reality is that simply physics (bandwidth is NOT the same thing as latency) still dictates local responses for highly interactive tasks. No doubt much of that will (and already is, of course) done in browser apps.

But to contend that everything will move within the browser is just as unsupportable as saying something like “all development will be done in language <insert your favorite language / framework here>”.

How many times has that prediction been made, in one form or another?

It’s just silly, really.

All the Same
I understand that true blue believers can take exception to everything I’ve said, except for one thing …. 60,000,000 apps in one month, on the best mobile web browsing platform ever … with some great tailored web apps (the newest google reader really is awesome).

Rather than arguing for what is effectively both the repeal of the laws of physics and universal world peace at precisely the same moment, perhaps it would be more productive to create more effective clients to use all that cloud-based services have to offer …

… and build these inside or outside the browser, as best fits the circumstances.

A quick shout-out to Reuven Cohen for noticing this particular post, which I’d overlooked in a category in my RSS reader that had over 755 unread items in it … today! Relatively ironic, wouldn’t you say?

Over at the cloud computing group in googlegroups there is an interesting discussion about optimal load-utilization. Along the way Tim Freeman brought up an interesting point:

Are there hidden costs at running this high in the first place? We’ve heard the opinion from someone who is in charge of buying 100s-1000s of computers a year that commodity hardware isn’t made to run at this capacity. That you’re not getting as much value for your money over time because of far higher failure rates (i.e., that failures don’t increase linearly with utilization and that there is usually a sweet spot)

So that got me to thinking …

Heat Really Does Kill
Obviously there are many factors in failure rates of computing equipment (spinning or simply processors etc.), but assuming that you have not-horrible power cleanliness the #1 enemy will be heat.

Heat. Heat. Heat.

So, with that in mind one important way stuff becomes server-grade (i.e., expensive, non-commodity gear) is to get better at cooling than commodity gear. Interestingly, server-grade stuff also tends to try to get that last hard-to-obtain chunk of performance out of the components as well as provide varying amounts of built-in redundancy, both of which exacerbate the heat problems considerably, causing the heat-dissipation to get even better, which requires more power, etc.

So in that sense what Tim’s contact has a point, since when running at full utilization most processors throw off lots of extra heat, necessitating (at the very least) extra gear to handle.

And there’s always the chance that the heat will be poorly dissipated, thereby resulting in increased failures … yet that does not mean that buying server-grade gear is the right way to go anymore. Far from it.

A Better Choice
A couple of choices come immediately to mind -

• use lower-power components (as in laptop grade stuff). These will naturally generate less heat, and thereby tend to reduce their self-inflicted failure tendencies.
• run much leaner power supplies than most folks want to supply off the shelf

There’s other ideas – some interesting, some dumb – but those are a few for starters.

Is the Commodity Gear Today What We Need?
Interestingly enough, most of the stuff that folks have bought to build out grids has been server-grade in drag, more or less. Just look at the components and the power supplies – high energy consumption processors, big power supplies, beaucoup fans etc. Not always, of course, but that has generally been the norm.

In fact, it’s this “server in disguise” gear that passes for commodity in most enterprise data centers today … fine so far as it goes. As Cameron pointed out in the thread you can run the current commodity gear at 100% utilization with no particular increase in failure rates. True enough, but what if we think more aggressively?

In fact, let me go so far as to suggest that if we really are able to run at 100% for months without a failure, then we’ve massively overbuilt the “commodity” gear.

Back to what will be possible in changing our infrastructure as we make our transition to clouds – public or private.

This is the Key – Absolutomente Crucial!
Underlying all of the power / failure related infrastructure choices is an unspoken reality – the real key to using commodity at scale is to ensure that the application will survive the failure of individual computers / drives / switches / whatever without losing a darn thing.

Once you do that, at the application level, then you are free to experiment with different infrastructure choices to your hearts content, different utilization rates, whatever comes to mind – provided that your apps don’t care.

In other words, many of the benefits that may result from cloud computing – flexibility, scalability, lower costs, reliability, and so on – are actually enabled at the application layer.

One more thing – when failure of individual computers doesn’t matter to the application then you can pick lower power stuff that is also very cheap – now you’re starting to talk about a great cloud infrastructure.

The New Black Commodity
So as you carry this thinking further then you can start to imagine a much more aggressive type of commodity, one as yet unrealized.

Start thinking of bare-bones, fairly dense components that are uber-cheap … sort of a lego-block approach. Cheap as in $300 -$400 cheap all-up. Perfectly suited for enterprise-grade clouds – public or private – at least those that play by these new rules.

There was (and will continue to be) quite a bit more conversation on this point – it’s one of the more interesting parts of commoditization. In any case, in a future post I’ll outline some more thoughts on the “new commodity” that I believe is fundamentally possible.

Ran across a funny pair of dialogues (from a Niraj J, obviously an app guy!) before and after the adoption of private clouds in a hypothetical enterprise. At least it would be funny if the before case wasn’t so painfully true.

Before (edited for clarity):

Applications VP – I need to unarchive some 300 GB of data and then use it for some analytics that I need to perform at least once every month.

Infrastructure Guy – 1GB costs about X $ and 1 LPAR with 2 CPUS is about Y$ per year. You need to multiply this by 5 years to get the ROI calculation for your project.

Applications VP – Wow! Why is the cost 1.7 times Frys?

Infrastructure Guy- Well it is all the overhead – The Company needs to pay guys like us who ensure that additional storage is installed correctly and that your group adheres to all the norms we have established.

Applications VP- OK (whatever … since I do not have any options!), when can I get it?

Infrastructure Guy- it will take 2-4 weeks after the purchase order is approved and quote submitted.

My only comment here is that it’s probably more like 8-whenever weeks in most enterprise shops, not 2-4.

In any case, here’s after:

Applications VP – I need to unarchive some 300 GB of data and then use it for some analytics that I need to perform at least once every month.

Infrastructure Guy – Here you go , call this API for Adding Storage and launching an instance. You will be charged by the hour.

Applications VP- Cool, I am charged ½ of what you guys charged me earlier and I have the ability to turn off my meter when I’m done.

Infrastructure Guy – Yes, They have cut down our group, and all my buddies who did not have scripting skills have been asked to go. I guess our overhead is now 1.1 X as compared to 1.7X. Besides if you consider the savings you get by switching your computing off when not needing it , we are probably cheaper than Frys.

On a serious note, this sort of on-demand flexibility will be just as appealing within an enterprise as it clearly has been outside.

Furthermore, I think it’s likely to be just as appealing to the operations folks as it obviously will be the applications groups, for two halves of exactly the same reason – it makes life simpler.

It’s just a matter of time.

Note that I’m not commenting here on whether the private cloud is built entirely within an enterprise, is provisioned on-demand from an external provider, or both. All of those are possible, and perhaps even likely. This is just to illustrate why people will – and do – care.

Structure08 is well underway and going well. A couple of good places to follow the flow, including Summize.

Also, check out my twittering here, and Sam’s here.

This time last year mobile devices seemed really stale … it had been years since meaningful change. There were lots of promises and rumors of good stuff to come, yet actual new devices all seemed to be hardly any different at all from the tired ones you held in your hand.

Small improvements here and there, but the big problems remained. Then the iPhone became the first of the new mobile platforms to become more real … and in the past year it has crashed into the mobile markets, becoming the first device to be a real enabler for what innovation in clouds and cloud-based apps are beginning to deliver.

The ability of these new devices to meaningfully participate in general web-based stuff is what sets them apart.

But that was not so obvious a year ago …

Forced to Change
About that time I left my old reliable Treo 700 in the seat pocket in front of me … minutes after landing from a red-eye to Europe, groggy and incoherent, unable to go back and retrieve the phone because of the (well-posted) rules in Lisbon.

This turned out to be about a week before the iPhone intro. I’d promised to all that I wouldn’t just run out and get an iPhone, but what could I do?

What I actually did do was buy the current Treo (750p) and the iPhone, and did my own mini-bakeoff, with the loser to be returned ignominiously. I’d been a Palm OS user for about seven or eight years (and two billion Palm corporate reincarnations), and been part of it going from responsive, stable, reliable hand-held goodness to a frustrating steam-driven, unable to handle threads handheld time bomb that regularly crashed during phone calls … ARGGGHHHH.

Tried but couldn’t get over the WM6 UI – yes it had a stable kernel, but every time I’d tried a WM device it seemed like interaction had been governed by the “5th People’s Directorate of Handheld Design”.

Pretty much the same for the Blackberries that I’d tried, though they’ve definitely evolved fast, and hung on to that addicting email / texting / now twittering title. Still no substitute for real keys for that stuff.

Yet, in the end my two week trial turned into two days – the Treo went back into hiding, and I hung on to the iPhone.

Why I Picked the iPhone Last Year
In the end it was easy – the iPhone was simply the first handheld that really did a decent job at web stuff. Particularly great for keeping up with RSS feeds, blogs, etc.

It was decent at email, contacts, texting, twittering etc – decent, but not great.

But the ability to take care of a bunch of web tasks far more than made the switch compelling for me. The wifi took the sting out of the slow mobile data connection, which helped. Besides, it never crashed during a call … not once. The stability alone was worth something.

Probably best of all, it turned out to be surprising durable. I think I dropped it hard about a gazillion times, and it kept working fine. Still does.

What has Bugged Me All Year Long
Lack of cut & paste is just stupid – still is, for that matter – but beyond that it’s been lack of

1. an SDK
2. active-sync
3. 3g

Well, all but the cut & paste are fixed with the iPhone 3G, with some cool bonus stuff as well (gps, ostensibly improved batter life, better initial price).

On the shaky list remains Apple’s irritating reluctance to approve developers and AT&T’s handling of data / voice plans.

Reaction to the iPhone 3G
Michael Arrington is lavish, Alex Miller is funny, Om Malik is well-reasoned, many analysts are ga-ga. Highlights include:

I Am A Member Of The Cult Of iPhone (Arrington)

I heard Apple’s new iPhone can sense your thoughts … I heard the new iPhone can change diapers and release a lemony fresh scent … I heard the new iPhone eliminates entropy (tweets by Alex Miller during the intro)

it’s time to shift attention to the most important question about this device: How much money will it make for Apple and its carrier partners? (Om Malik)

look for units sales of 14 million in calendar 2008 and 24 million in calendar 2009 (Mike Abransky, RBC Capital)

My Take
I think the bottom line likely reality is really pretty simple – Apple is going to sell about a billion iPhones, and with the release of the 3G the platform itself is reaching critical mass as a meaningful mobile device.

While it is by no means even vaguely close to perfect, it’s the first of the next-generation devices that seemed for so long to have been stuck in a maddeningly crazy “coming soon” time warp … to move out of that time warp and actually become real. And it is very real.

Android could be next, with everyone else fighting it out to either stay relevant or perhaps also gain relevance. Btw, here’s a shout out to CrazyBob and the rest of the Android team for forcing Apple to accelerate the SDK release … looking forward to seeing production Android devices in the mix.

I think this the iphone is the first device to begin to match what’s happening in the web – particularly clouds & cloud-architected apps – enabling fundamentally new stuff… particularly with the SDK. Of course, Apple could still screw this up, but I hope not.

Besides, it’s actually not too bad for old-school phone calls!

Self-healing is at the core of much of what makes application fabrics work – whether you deploy on commodity in your own place or on a cloud.

It just has to be a non-event when stuff breaks – both for the work in progress, and for the “structural integrity” of the app fabric itself. Ensuring both enables the app fabric to provide the simple abstraction of a reliable, uber-scalable computing surface.

At any rate, I was reminded of this a couple of weeks ago when this video first made the rounds… cool stuff.

I particularly like it because it’s a great illustration of the value of simple goals & simple organizational rules, both in theory & in practice.

Take a few minutes to watch it, if you have not already … and ponder how cool it is when application developers can just assume that their underlying infrastructure behaves as resiliently as do these robot blocks.

Enjoy.

The video is courtesy of New Scientist, and the work is being done by Mark Yim and company, of modlab (the Modular Robotics Lab), at the University of Pennsylvania.

Great work guys … keep it up!

I’ll admit that when MarkSu first blogged about building a scalable Twitter / Twitter client about a year ago (he kept going with his second, third, and fourth posts) I just started laughing, because like Stephen Forte I just couldn’t figure why anyone would care.

Then I started looking at it closer, drawn primarily out of curiosity about why they kept going whump. Of course, most of the early adopters are fairly vocal, so the commotion over these failures was … well is high-volume.

After using twitter for a couple of weeks I think I mostly agree with Alex Miller on why he likes twittering.

I think the “rules of engagement” will morph quite a bit as usage grows, and I do think usage will grow. Only, like a growing number of folks I’m not sure there’s any particular need for a centralized service, particularly one that isn’t stable.

Maybe Twitter will fix it’s scaling problems (very doable) and discover a revenue model, or maybe the whole idea of micro-blogging will go decentralized, or maybe some combination of both (which gets my vote for most probably outcome right now).

So MarkSu, I have to admit: you were right!

MarkSu also just bought a SmartCar (actually two – one that’s pictured above and obviously already here, one on the way), but I really doubt that he’ll get me to do one of those – but you never know!

I read Robert Scoble’s post on “noise in the net” with morbid curiousity today. It’s a pretty interesting post, really. Seriously.

Oh, the glorious noise! Everyone loves beating me up for causing the noise. No, I am not the cause. I pass it along. You should see my inbound streams. Every second or two a new Twitter is aimed at me. Every few seconds, a new blog post comes into Google Reader. Every few seconds, a new thing on FriendFeed.

Scoble then goes on to give a few ideas on how to get past the noise and make some sense out of it all. The suggestions are OK, but not really great. They’re really more thoughtlets than anything useful.

Irony on ‘Roids
Anyhow, do you want to know something really ironic? I was just about to remove him from my twitter “follow” list, precisely because the amount of noise and gunk that he’s been generating has been enormous, and the value has been … well, not enormous.

Sort of like the opposite of Dolby Noise Reduction … but maybe worse. He not only passes along the noise, he adds more to it and injects back into the same stream, making it go from meta-stable to insanity inducing.

Sort of a human positive feedback loop. As in this excerpt (from the linked article)

A system in equilibrium in which there is positive feedback to any change in its current state is said to be in an unstable equilibrium … the end result of a positive feedback is often amplifying and “explosive”, i.e. a small perturbation results in big changes.

Uggh.

I know we’re trying to figure out how this can all work, be helpful and all that, BUT … is that really a good model for social communication? Then out of all that gunk he comes up with something interesting.

Maybe I’ll wait until tomorrow to break the loop…