Slashdot Mirror
Open Source Facing a Difficult Battle For Cloud Relevance
A recent eulogy for open source's relevance to cloud computing by Redmonk analyst Stephen O'Grady caught the attention of Matt Asay, who breaks down the difficulty of this David and Goliath problem. "In a world where horsepower matters more than the software feeding those 'horses,' in terms of the entry cost to compete, and where big vendors like Amazon and Google are already divvying up the market, the odds of a small-fry, open-source start-up challenging 'Goliath' are slim. It's not a new argument: Nick Carr has been suggesting for some time that only a few, big companies can afford relevance in this hardware-intensive business. Given this fact, O'Grady thinks the best we can hope for (and he thinks it's pretty important) is 'a loose coalition or confederation of [open-source] projects and vendors that will together comprise an increasingly viable top to bottom alternative to some of the cloud providers today.' He includes projects like Puppet (Reductive Labs) and Hadoop in this mix, but is careful to point out that he doesn't see a full-fledged, open-source alternative seriously challenging the closed platforms of Google, Amazon, Salesforce, and the other mega-clouds."
But if open-source can hit the bullseye, the rest of the dominoes should fall like a house of cards. Checkmate.
Really, is this situation THAT MUCH different from what we have today?? What are the chances of a small mom-and-pop start up create a virtual bookstore to rival Amazon, or an Internet services infrastucture empire to rival Google??
should be "Cloud computing facing a difficult battle for Relevance"
--
Stay tuned for some shock and awe coming right up after this messages!
Aren't those vendors built on top of open source? If I remember correctly, Google uses their own Linux distribution, Amazon uses redhat, and I have no clue what salesforce uses but I imagine that it's probably some form of open source OS since they can save a lot of time and money using that instead of Windows when we're talking thousands of servers. The cloud revolution, if anything, was brought on my open source since it's made deploying thousands of servers cheap and easy. If the companies had to pay for licensing of software on all of those servers or roll their own OS, they would have built up (buying fewer, more powerful servers) rather than building out.
Now you come to mention it, we do already have all that in Windows.
Professor Karmadillo Songs of Science
Maybe I'm missing something big here, but isn't "cloud computing" largely just a data delivery service, and not really "software"? It's kind of hard to get a handle on "cloud computing" since it's such an amorphous buzzword. Can someone give me a real example of an application that's "cloud computing" based. I thought my little weather app telling me the temperature might be defined as "cloud computing".
If the above is true, I don't see how OSS can really make some big impact on "cloud computing" any more than it can make it on websites. If it's not true, how could OSS big a big player in "cloud computing"?
AccountKiller
Come on people, most of Google's and Amazon's could are run by Linux / BSD with costume modifications to adapt to the task at hand.
If the article would state that these companies are not giving back much to the community in relation to what they take, then yes, that's probably true but they still rely heavily on OSS software.
For me the whole article completely misses the point, but maybe I'm missing something here.
Also: cloud computing is not going to take over everything. It is useful for certain situations like massive indexing, data backup storage and some forms of HPC (though the last group mostly build their own data centres or rely on distributed computing). The everyday business will not participate much.
It's really hard to see how free software isn't relevant to "cloud computing" services when you can basically build your own using them. Apache/MySQL/Php can let you build quite a bit...maybe that's not enough to be cloud certifiable er...certified but it works for me.
The other issue here is market leadership and time-to-market. Admittedly this speed is somewhat lacking the free software world because the motivations are different but in the long run, free software will win out as it allows more of the best minds to collaborate to build better systems. I'm looking forward to a user/customer owned coop cloud solution and perhaps another one that consists of ready-to-download virtual machines that I can run on my own hardware wherever it may be. A project called Eucalyptus is a step in the right direction in this space.
Some of these network services are starting with the right ethics in mind and it's those we should be talking up. With identi.ca, libre.fm, Eucalyptus and other projects making progress each day, free software(not open source) is anything but dead.
That is called FUD. Get the current buzzword related to technology, and just go and make like the OO movement is off the hook on it.
--- "When you gotta do something wrong. You gotta do it right. (Fighter)"
I think cloud computing will be forever stuck in the realm of casual consumers and enterprise. Someone who purchases a $100-$400 netbook and browses the web will probably be the primary demographic here. And for large businesses, they'll have their own enterprise-wide cloud computing solution. This is really just a web-savvy interface on top of the traditional mainframe infrastructure. For those of us who have been computing for some time now, or require absolute control over our privacy and security, we'll stick to the traditional modus operandi of desktops and laptops.
I really don't understand the long-term value proposition of running your stuff on a public cloud. I can, however, see the IT cost advantages of a properly automated internally managed cloud for internal IT needs. You can get more efficient utilization of hardware and easier administration using virtual servers in a cloud configuration. Of course, there are open source solutions for that, so I'm not sure where the notion that open source can't compete in this area is coming from. Hell, many of the software solutions for this sort of thing are based on the open source Xen these days.
"Cloud" has been, in many venues, too narrowly defined as being "outsourcing to someone else's cloud", when in fact if you already have an IT department that already manages your servers in house, you can probably get more bang for your buck building your own cloud and converting your existing servers to virtual machines running on it.
It's also incredibly dangerous to say the amount of horsepower you have is the most important thing for cloud computing. The most important part of the cloud is the automation and management software. If either of those two things are inadequate, the cloud will be inadequate and very expensive to maintain. The software is the key to a successful cloud implementation. The end result of a successful cloud implementation should be more efficient use of hardware and more efficient and easier administration, resulting in an overall reduction in cost. If the software pieces aren't in place, you won't reach those goals.
'In a world where horsepower matters more than the software feeding those "horses"...'
Wrong already! Software does the work - the "what" of solving problems. Hardware, while of course necessary too, is basically a fungible commodity - the "how". To use a counter-intuitive but revealing analogy, software is like the car while hardware plays the role of fuel.
Good software is still fairly rare, whereas state-of-the-market hardware can be cheaply and plentifully obtained from several alternative sources. So the article has it exactly the wrong way round: it's software that is important, and hardware that plays the supporting role.
I am sure that there are many other solipsists out there.
I'd suggest that they are likely to grow to being an important part of computing, but no bigger than, for example, the large-server-and-Oracle part. (full disclosure: I'm a capacity planner, so most of my income comes from just that part).
The disadvantage is that my cost per transaction is greater than if I had a steady load and ran my own machine room. The fees I and the other customers pay a cloud service have to cover their whole machine room, whether it's it's busy or not, plus their profit.
So I see a natural evolution for a growing business. While they're small, they'll build a LAMP or Java stack on a small machine in the back room. If they grow slowly and steadily, they'll buy more, larger machines for the back room. If they grow without bound, they'll jump to LAMP-on-cloud or Java-on-a-cloud, with a few code changes as possible.
Once they have mastered that, they'll move back and forth, depending on the business growth rate. If they grow too fast, they'll do a lot in the cloud. If they grow slowly, they'll have a cloud presence, but try to process as much in their own machine room as they can, to improve the profit margins, using the cloud for overflow and to run during my machine-room upgrade.
Conclusion? common software between the cloud and the machine-room is important. Look for any standards developing in the LAMP/SAMP space, like the DMTF incubator at http://www.dmtf.org/about/cloud-incubator Look for Java offerings for business, like http://blogs.sun.com/cloud/entry/communityone_cloud_recap When you're there, specifically look for virtual machines that will run in the cloud. Finally, look for load-balancing mechanisms that will send your work to two different places, under your control, sometimes called "application distributors".
Don't assume open source is at a disadvantage: if you can run your stack on a free VM on a standard-conforming cloud, however commercial it might be, then your computing can remain free of the control of others.
--dave
davecb@spamcop.net
This is wholly misleading.
Salesforce is crap, there are more competitive alternatives and most people avoid it like the plague. See Siebel for an easy example. Not to mention most people dont' want to have to a: rely on salesforce or b: give up the control that enterprise can and should have.
This is one reason cloud as a concept fails: lack of enterprise control. It has minimal enterprise interest for this reason. Also add a lack of legal certainty as to apps hosted in the cloud and you have something most corporations will not touch with a 100 ft pole, let alone a 20 footer.
Cloud computing is in essence one step worse than proprietary software, in that not only is your data locked up in proprietary formats but it's now hosted on someone else's servers too, making you even more dependent on the service provider.
On the other hand, unlike software, they are providing a service with contracts guarantees... I would demand a guarantee of a certain level of uptime, and a guarantee that i can always take my data out in a standard format if i want/need to. Very few proprietary software guarantees you the ability to retain your data in a standard format that can be imported into a competing product or service.
http://spamdecoy.net - free throwaway anonymous email - avoid spam!
Could Computing is simply a service provided over the Internet that is scalable and virtualized.
In short the software is in the web browser, while the data is stored somewhere else like on the servers. The word "Cloud" is a metaphor for the Internet.
This is not just an ordinary web application, it usually involves a virtual machine of some sort so that the web applications acts like a desktop application within the web browser. One that can be scaled to handle an almost unlimited amount of users.
So for example the PHPBB2 Forum software is a web application, but not a Cloud Computing application. Google apps, on the other hand works via a virtual machine and software as a service so it qualifies for cloud computing applications. Google apps do GMail, Word Processing, Spreadsheet, etc in the web browser under a virtual machine but the data is stored on Google's servers.
The reason why open source developers don't support cloud computing is because they feel that it locks the users into third party technology and exposes their data across the Internet in violation of privacy that others could spy on it or capture it via packet sniffers. So OSS developers try to avoid making cloud computing applications as a matter of personal ethics, etc.
Remember, Slashdot does not have a -1 disagree moderation, and no, troll, flamebait, and overrated are not substitutes.
Is Nick Carr just some academician who spins crazy theories just to get attention, and maybe make some money?
He seem almost like a professional troll, with sensationalist, often inflammatory, subject lines like "is google making us stupid."
Is there any reason to assume that Nick Carr knows any more about the future of IT than the average bum on the street? Okay, he's educated, since when have whack-job college educated predictors ever proven to be more accurate than flipping a coin?
As with anything, it entirely depends on who you ask.
'Scalable' does seem to be nearly ubiquitous for the concept of what 'cloud computing' means. Virtualization is common, but not a prerequisite.
Your description seems to indicate that a 'virtual machine' in this context is referring to the more application-style of what runs in the browser behaving like an application. By and large, this style of making more extensive use of javascript to give a more 'desktop' feel to web applications is a mark of the 'Web 2.0' buzzword (though the context most widely credited with coining the phrase didn't speak to that at all). When people talk about virtualization in the cloud, they almost always refer to OS instances being executed with a virtualization layer abstracting them from the real hardware (and making some of the more fatal hardware situations appear more like a simple reboot to the os instance, and other imminent failures no problem at all). Some rely on higher-order application-level redundancy, and forgo the virtualization aspects (many of the IO intensive workloads are still very reluctant to embrace virtualization, for one). Others even rely on 'user-level' redundancy (i.e. user sees a problem, hits refresh).
Some think of a cloud as a computing resource in which the usage picture is highly dynamic without strict mappings to where things must happen.
Some think of Cloud as a sort of spiritual successor to 'Thin Client', often extended to the internet. Where Thin clients were almost universally thought of as essentially remote displays, the reinvention in the cloud context generally has a more sophisticated client that is fed data to interpret and manipulate, though it's nearly required that client-side data persistence not be a critical pre-requisite. A total destruction of a 'client' in this definition of cloud has little more permanent consequence than 'thin clients'. I.e., Valve's Steam, where you could throw your computer off the top of a building and theoretically recover all your purchases, and, for the games that support it, the settings you use. In steam, the coupling between client and 'cloud' is relatively loose (some aspects can operate completely offline, and save-games may not fit the definition) , whereas 'google apps' is relatively tight.
phpBB could be considered a 'cloud' application, so could BBSes, so could a lot of things if they came to popularity *right now* instead of when they did. Essentially, most all webapss meet *someone's* definition of cloud, and it's such a vague term with no authority behind it, no one can call them wrong for the most part.
I don't think OSS developers avoid making cloud applications no more than anything else. The actual code behind many cloud computing implementations is OSS (Hadoop for one), but people refer not to the software, but to the popular sites that use the software. OSS is a phenomenon built entirely around how software is designed and produced. By most all definitions of cloud computing, it is a phenomenon that is built entirely around how software is put into implementation, usually with the characteristics that the users don't even know what software they are really using.
As far as OSS goes, cloud computing might actually be easier in that environment. The companies know the value lies in the data being managed moreso than the software used to manage it, and will risk others leveraging more for the sake of outsourcing development costs to a community. However, the philosophy behind OSS, as you say, may naturally lead some to worry about control of their data.
XML is like violence. If it doesn't solve the problem, use more.
As far as I can tell, open source and Linux are being used far more widely in cloud computing than in corporate America. Cloud computing is going to be a cut-throat business, and it will be tough for companies like Microsoft to compete. Few of their usual dirty tricks work. And the cost of switching is low.