Slashdot Mirror
Is OpenStack the New Linux?
snydeq writes "As the self-proclaimed 'cloud OS for the datacenter,' OpenStack is fast becoming one of the more intriguing movements in open source — complete with lofty ambitions, community in-fighting, and commercial appeal. But questions remain whether this project can reach its potential of becoming the new Linux. 'The allure of OpenStack is clear: Like Linux, OpenStack aims to provide a kernel around which all kinds of software vendors can build businesses. But with OpenStack, we're talking multiple projects to provide agile cloud management of compute, storage, and networking resources across the data center — plus authentication, self-service, resource monitoring, and a slew of other projects. It's hugely ambitious, perhaps the most far-reaching open source project ever, although still at a very early stage. ... Clearly, the sky-high aspirations of OpenStack both fuel its outrageous momentum and incur the risk of overreach and collapse, as it incites all manner of competition. The promise is big, but the success of OpenStack is by no means assured.'"
...projects to provide agile cloud management...
Whenever I see "blaw...blaw AGILE blaw...blaw", I stop reading.
If you want news from today, you have to come back tomorrow.
I've HURD this before.
As a general rule, the only way to build something large and complex that works is to grow it from something small and simple that works.
OpenStack is a Linux distribution organized for deploying a compute cloud. Linux is the new Linux?
It sounds very exciting until you look at the code. Then you realise that the quality in the project is entirely in the marketing, and there's nothing of worth code-wise at all.
And something that does everything, no less. In general, this means it does nothing well. Big egos are just the hallmark of failure. Lets see whether anybody even remembers this in 20 years. Personally, I doubt it.
Most ACs are not even worth the keystrokes to insult them. Be generically insulted by this and ignored otherwise.
Just applying the rule I saw somewhere: if the title is a question, the answer is "No".
May Peace Prevail On Earth
Whenever I see "Whenever I see",
Insert self-referential sig here.
Is there some deficiency in Linux and the various BSDs that OpenStack is intended to remedy?
There must have been a lot of development going on in the OpenStack camp during the past year. Last time I checked many features that were already available in other open source cloud platform products were work-in-progress and setting up and configuring a functional private cloud was cumbersome at best. I wonder how they have managed to gain such publicity and backing over more mature competitors.
Just testing out that question in the title thing... :)
Hey, Windows users, there is no such thing as "forward" slash, there is only slash and backslash.
Just like Linux has been the new DOS.
No way to compare pears and beans.
Sent as ripples into the electromagnetic field. No single photon has been harmed in the process.
I was misled by the summary. This isn't a whole new OS from the ground up - it's a Linux distro with some python code included.
Is a Linux distro with some python the new Linux? Umm... yeah... how about no.
FTFA:
It's a big change of scene for me. My last full-time job was in Windows Server product marketing, which prevented me from writing for InfoWorld or anyone else except my Redmond bosses for four years. Now I'm back in the game, laptop battery fully charged, ready to chronicle the next big thing in open source.
It sounds like marketing because it is marketing.
I don't believe it's a Linux distribution because it's intended to be installed on Fedora, RHEL, Debian, Ubuntu, etc.
http://wiki.openstack.org/InstallInstructions/Nova
As a "cloud OS for the datacenter" Open Stack clearly has to iconoclast on empowering croud-sorced segregation-effects within the namespace of its initial synergies. Anything else would be a paradigm shift.
Oh, the beautiful gloss of greality!
To me, the word "cloud" describes a type of service, but most marketing uses it to describe anything that involves the network.
It's meant to be syllogistic.
As in:
Linux:Operating Systems::OpenStack:Cloud
At this point, though, OpenStack is still pre-1.0, perhaps equivalent to Linux circa 1993. Whether it can polish up and continue to deliver what is needed is yet to be seen.
The impetus behind cloud right now means that this will be a lot more high profile than Linux was in 1993. There's all sorts of politics (eg Why Citrix Left Openstack) at play, and no one has an OpenStack cloud of any significant size running. OpenStack has been tooting its horn for 18+ months and yet the most advanced player is really just going into production. Rackspace clearly sees OpenStack as an avenue to leverage outside development in an effort to go after Amazon, but whether that makes it viable for other people - and thus creates a rewarding ecosystem - has yet to be seen.
Any time the word "cloud' is used, a real tech wants to punch you in the face. "The cloud" is a title so we don't have explain how the internet works to moronic Mr.CEO and pals.
Amen, brother. It's really getting annoying to me lately.
There's a term used called "cloudwashing" that covers inappropriate use of the term cloud, but cloud technology is real and every company in tech is pouring money into this transition.
Anyone who has worked in IT in large enterprise has seen the benefits of virtualization in action; there's an enormous amount of capex and opex savings, and VMware basically dominates the market. There's a reason 99%+ of the Fortune 500 have an ELA with them.
The same principles behind that revolution are now reaching into the public space, and looking to blend the private IT compute farms with public cloud resources as well; plus more apps being deployed as SaaS, and more apps being developed on PaaS stacks; all the technology of big data (eg, Mongo), messaging (eg RabbitMQ), and so on just form a virtuous circle with this trend. Apps become more able to run in generic clouds without requiring very specific hardware control, and thus IaaS clouds become more attractive.
If you're in system, network, storage, or security administration, or IT of any sort, and you're not learning about this, you're basically a COBOL programmer waiting to be put out to farm.
We have a stupid research group that is always chasing after the latest trend. "The cloud" is their new shit and they want OpenStack bad. They don't know why they want it, they just do. Of course when our Linux guy sets it up for them, they can't use it because they have no idea how. They don't like the idea of just using VMWare for some reason. It isn't cool enough to them.
There's a term used called "cloudwashing" that covers inappropriate use of the term cloud, but cloud technology is real and every company in tech is pouring money into this transition.
Anyone who has worked in IT in large enterprise has seen the benefits of virtualization in action; there's an enormous amount of capex and opex savings, and VMware basically dominates the market. There's a reason 99%+ of the Fortune 500 have an ELA with them.
The same principles behind that revolution are now reaching into the public space, and looking to blend the private IT compute farms with public cloud resources as well; plus more apps being deployed as SaaS, and more apps being developed on PaaS stacks; all the technology of big data (eg, Mongo), messaging (eg RabbitMQ), and so on just form a virtuous circle with this trend. Apps become more able to run in generic clouds without requiring very specific hardware control, and thus IaaS clouds become more attractive.
If you're in system, network, storage, or security administration, or IT of any sort, and you're not learning about this, you're basically a COBOL programmer waiting to be put out to farm.
Funny, we just hired two COBOL programmers at $80K each to maintain some legacy mainframe systems. When cloud technology can permit hard core data entry, say for insurance records or the like, then I'll worry. But until then, throughput is more important than an app being able to run from wherever in the cloud. Besides, in my line of business. We don't run apps. We run programs that process millions of secure transactions. We have data entry clerks that key documents and data that can't be captured electronically.
You would probably say that we have our own private cloud. I would say that we have our own methods to allow secure access to our internal systems. By the way, I would predict that there will be COBOL programmers still programming even after cloud computing has been replaced with the next marketing hyped phrase.
As a general rule, the only way to build something large and complex that works is to grow it from something small and simple that works.
Which is why the Saturn V booster used in the Apollo 11 mission to the moon was built out of Legos.
-- Terry
I stopped reading this thread when I saw "blaw .. blaw AGI
This reminds me of a guy I knew in ~1994, who was griping that all his experience was in COBOL, and after getting laid off from making $75k/year, he couldn't find another job. At the time, I was in college, and so I wasn't really familiar with the idea of keeping your skills updated...
So I don't know that I would recommend cloud for you; there are reasons to use it, and reasons not to use it. As the technology and ops experience matures, it will be easier to adopt - basically like any tech. But for almost everyone, there are real benefits. Both capex and opex; and some people are using cloud in a way that their capex savings is ~0 (or negative) but their opex savings is huge. (See: Netflix running their entire infrastructure with 3 admins) Program ~= App. I file my expenses through an Oracle app, that runs in a cloud, that automatically fetches corporate card transactions from Visa, and lets me roll them into an expense report.
I'm one of the authors of Securing the Virtual Environment, and my co-author is a QSA, and one of the points of writing the book was to talk about the fact that cloud *can* be secure and can be compliant. (Although in the case of a public cloud, obviously compliance requires underlying compliance by your provider, as well as your own processes) Of course, there are a bunch of risks, too - but there are, for example, cloud services that have passed HIPAA and FISMA audits.
In short, cloud is more than just a buzzword; it's an evolution in the technology that powers IT. I'd say it's more evolution than revolution, but it is more than a buzzword.
...we will hear every year for the next 20 how this year is the year of OpenStack on the desktop?
Dog is my co-pilot.
"the New Linux" is an analogy (you literal dorks). From the FAQ http://openstack.org/projects/openstack-faq/
What does it mean for the cloud ecosystem?
This is not yet code that comes with certification from operating system or hardware vendors. Instead it's aimed at providers, institutions, and enterprises with highly technical operations teams that have the capabilities and needs to turn physical hardware into large-scale cloud deployments.
Still, wide adoption of an open-source, open-standards cloud should be huge for everyone. It means customers won't have to fear lock-in and technology companies can participate in a growing market that spans cloud providers.
A great analogy comes from the early days of the Internet: the transition away from fractured, proprietary flavors of UNIX toward open-source Linux. An open cloud stands to provide the same benefits for large-scale cloud computing that the Linux standard provided inside the server.
tomorrow who's gonna fuss
Considering that a full Debian system (binaries for one platform) comes on eight DVDs, I think this project, whatever it might be, has a long way to go before it can really claim to be the "most far-reaching open source project." More mainstream, perhaps, but far less ambitious.
I always think of Betteridge's Law of Headlines
Headline I'd like to see: "Is Betteridge's Law of Headlines actually correct?"
"Slashdot - News and Chat Sites Deviant". (Click "homepage" link above for details).
Yes. Linus Torvalds is well known for his huge ego. ... or was it his self-deprecating humor?
Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
God I wish I had mod points. I'm reading the comments and wondering how so many people could ignore the giant whooshing sound of the point going over their heads.
I work for a .au university on an openstack cloud service providing compute infrastructure as a service for researchers. we currently have about 2000 cores on 84 compute nodes (plus swift storage and volume storage and nova-api and database and so on) and about 700 users (and anyone with a login at any australian university can have a login with a small allocation of cores/memory/cpu-time, with larger allocations on request), with another 2000 cores ready to go as soon as the regions/zones/cells (or whatever they're calling it this week) code in nova works. we're currently running a very hacked up version of Openstack "Diablo" on Ubuntu Lucid, and will be upgrading to Openstack "Essex" on Ubuntu Precise in the very near future (next week, most likely, unless something comes up to delay us).
any of our users can login, spin up a few VMs through the web dashboard (or via command-line tools), and run or develop the code for their research project. we've also got people working on tools to enable researchers to spin up entire HPC clusters on demand within our cloud.
the openstack "Cells" code is crucial to us, because over the next couple of years we'll be expanding the research cloud in stages to have nodes in more universities - it's a national science infrastructure project, funded by the .au federal government. in the not-too-distant future there will be tens of thousands of cores.
So, yeah, there are openstack deployments of significant size already running.
nothing's ever perfect - the current openstack code has a lot of problems....but there's also a lot that's good or even great about it, and the rate of improvement is impressive.
actually, one of the biggest problems with openstack is terminology. the various components of openstack began as separate projects, and they have re-used the same terminology (especially the word "zone") to mean very different things, which tends to make things confusing. that's one of the things that is being worked on.
So you are saying that the cloud is just an alias for virtualization? Given your vast studies in this arena, can you explain the difference between hypercloudification and paracloudification?
Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
Linux already controls the life-cycle of VMs.
Guns don't kill people; Physics kills people! - John Lithgow as Dick Solomon on Third Rock From The Sun
I strongly disagree that it's a general rule.
I find that a lot of people I would not ordinarily view as idiots have this absurd idea ingrained in their psyche that it's possible to incrementally get from thing ABC to thing XYZ. Mostly I have to believe that these people have never had to reverse engineer anything.
One of the places this happens most often is in Open Source software, where people have drank the Eric Raymond kool aid about "The Cathedral and the Bazaar". Sometimes you need to build a cathedral, and there's simply no way to get there from a mud hut. I think SourceForge "declarative projects", where people believe that by simply declaring an Open Source project, they will get droves of willing bodies to implement it from scratch. This is the big lie that was Mozilla for a long time, until they finally had working code for people to tinker with, and even then, they attracted tinkers. It shows: they ended up with a bricolage.
Another example is the idea that you can get from a system with a small set of capabilities to a system with a large set of capabilities without sitting down, mapping out the problem space, and then designing a framework in which it's actually possible to represent the entirety of that space. This misconception is often perpetrated in things like Portage, which is a glorified package management system which is frequently pressed into service as a build system, a task to which it is demonstrably unsuited.
I'm generally annoyed when anyone portrays something that was revolutionary at its time as part of a natural evolutionary progression on a straight line route from point A to point B, and that it as somehow "obvious" that this revolutionary idea was the next step in the progression. 20/20 hindsight (or Monday morning quarterbacking, if you prefer) aside, it does a great disservice to the revolutionaries who came up with the idea/application at a time when no one else was doing anything similar.
-- Terry
Interesting. I'm only starting looking into cloud based systems and it strikes me that security is going to be THE big hurdle in making a valid business case and the fact that security seems to be based around policy compliance seems to be a chink in the armour (not that any connected service needs any more than already exist it appears). Looking at the Amazon description for your book though it seems I may be misinformed, I'll have to check it out :)
The only important thing about OpenStack is that it supports implementation of managed, dynamically allocated and partitioned clusters (what "cloud" really is) with LXC, a non-virtualization host partitioning technology.
Contrary to the popular belief, there indeed is no God.
It was until now.
Play Command HQ online
It sounds good on paper but actually getting OpenStack to work on our hardware was a nightmare. It took multiple man months just to get our test machines to fully boot up. We recently switched to Azure and we're not going back any time soon.
We look forward to reading your book.
Il n'y a pas de Planet B.
I always think of Betteridge's Law of Headlines
http://en.wikipedia.org/wiki/Betteridge's_Law_of_Headlines
Is Betteridge's Law of Headlines true? .... though there is only one logically consistent answer.
Only you can decide!
Is Betteridge's Law of Headlines false? I just realised that I could do better. Now I've got you ... answer that one!
A suite of tools to manage a data center being the "most far-reaching open source project ever"? Seriously?
What a joke.
Well, it seems indeed, you'd need a little bit more knowledge. Yes, there's "the cloud" (as in: stupid marketing journalistic word), and real IaaS. Openstack is the later, and there's nothing dirty here. The moron might not be the one you think.
Maybe if you tried to install Openstack on some hardware, you wouldn't write this.
OpenStack is the answer to CPanel or Plesk, nothing more.
They have absolutely nothing to share. If you are saying this because of the Horizon dashboard, then you are mistaking. IaaS has all to do with API, using it like you would with a web interface is stupid, in this case, you need VPSes, not cloud computing.
So where you might only need one highly-tuned VM, you need instead 10 poorly tuned VM's to get the same performance. This allows Amazon and other Cloud providers to charge you by the hour at 10 times the amount.
Exactly for what reason you would spend some time to highly-tune a VM not in the cloud, while you would when it is in a IaaS? It doesn't make sense. In both situations, you can skip the tuning if you are lazy and have enough money to buy computing power. Also, what allowed AWS to have such an expensive pricing model is just the fact that their was very few competitors, and Openstack is changing this.
One bad sector on any hard drive= all virtual machines lost.
WHAT??? Come on, you clearly didn't understand what it was about. On the cloud, you'd be storing stuff on highly redundant storage. Like, your images would be stored with Swift (with glance using it), meaning that you'd get at least 3 copies of all files. And your instances would be disposable (eg: if one crash, it's not a problem, just fire-up another one stored in Swift). So no reason to loose anything if only one HDD crashes...
So I don't understand why anyone would use Virtualization at all.
Indeed, you don't understand...
The only proper way to setup cloud infrastructure requires a much larger investment: 2 NAS+2VM hosts, so 4 physical machines.
Nobody pretended that you could start with very few investment. The model only works when you have the need for a lot. BTW, Openstack doesn't use NAS, it has nothing to do with it. The whole point of Openstack is to use "commodity" cheap hardware. Typically, you'd setup Swift WITHOUT RAID for example.
The rest of our hardware assumptions are just wrong (you're talking about iSCSI, NAS and so on, when, again, Openstack doesn't need them...).
Funny you should mention HP: https://www.hpcloud.com/ It looks like they are implementing OpenStack very similarly to Rackspace, which should help back up OpenStack's claims of freedom from vendor lock-in. -Dan
I did not mean to imply that the "cloud" is not important, or at least won't be. I was really commenting on the COBOL side of things. However, with the exception of the monetizing of consumer products, such as Netflix that you mention. Aren't most cloud applications really just an extension of the client server model that we were all using in the 90s (your expense account example for instance). Of course, back then, they were all running on TCP/IP on internal networks or across T-1 lines where as now they are running on TCP/IP across the internet.
It seems that what cloud computing really means, today, is the use of the public network to replace private networks that relayed the same information. Granted such a use has more capabilities, particularly being able to access from almost any where, but the underlying technologies aren't really different than 20 years ago (thin clients with data bases on the back end -- although the tools have improved since then).
Even in our mainframe environment, we use a lot of COBOL on the back end with Java on the front, therefore leveraging what each tool is best at. I would agree, that the cloud is more than a buzzword, but it is over-hyped by marketers and the media. It is just a natural progression (evolution, if you will) of what came before it, from these applications running across private networks to public networks. I do look forward to buying and reading your book.
So, the general attributes of cloud for IaaS, offhand, are:
- Elasticity; you can provision and deprovision it dynamically and rapidly, and you pay only for what you use (and granular billing to go with it)
- Redundancy "under the hood"; your specific instance may fail, but a cloud service should heal without intervention from a tenant, beyond doing things required by their instance(s) restarting
- Multi-tenancy - meaning many unrelated entities can safely share the same hardware with a separation of concerns
- API interfaces
- Accessibility over a network
Some people would include a lot of other attributes, such as "linearly scalable" (ie, 1 instance = N units of processing, then 2 instances should = 2N units of processing).
Ultimately, the promise of cloud computing is to deliver just as much computing as you need, where you need it, only for as long as you need it, with ~0 setup needed on your part. If you've ever provisioned servers, you end up asking something like:
- Do I need shared, dedicated servers, or my own colo space to set up?
- Do I need routers, firewalls, load balancers, vpn concentrators, etc?
- Which things need (for security/role reasons) separation?
And then, what's your timeline when any of those answers change?
Cloud handles application scale-up and scale-down more gracefully; this is one of those things that's been driving virtualization in the enterprise for a decade; enterprises can consolidate servers, and old applications can share a tiny slice of hardware but still not be end-of-lifed, rather than needing their own server to run on. Applications which have a sudden burst of popularity could conceivable scale up massively - imagine a world where no one is ever slashdotted.
Virtual networking can give every application its own isolated network with its own firewall policy, using ~30mhz worth of cpu.
Anyhow, this sort of thing has driven virtualization in the enterprise for a while because of capex cost. The average utilization of non-virtualized servers is, iirc, ~30%; post-virtualization, it's 80-90%+. That means enterprises that use virtualization simply spend less than half on servers and the costs of maintaining them. Then there's opex. Rather than the complex provisioning associated with sizing, installing, and maintaining bespoke computing for every user/org/BU, the IT process can be streamlined to having a one-size-fits all provisioning, and the virtualization/cloud layer can carve it up dynamically. You have way fewer people needed per piece of hardware. To say nothing of how the resource sharing makes it self-healing. Physical server dies? The virtual machines that were on it power up automatically on a different machine. (In fact, VMware FT can actually add an application-independent hot-failover to any x86 server; physical hardware dies, the shadow copy immediately resumes running with the full state on the failover hardware)
Thin clients hitting servers (ie, dumb X terms hitting mainframe-type servers) was a similar concept in the sense that you were time-sharing resources, but this makes a similar arrangement possible without operating system dependencies, with application portability (ie, I can move a virtual machine from one cloud at one provider, to another cloud at another provider, about as fast as the bits from the virtual disk can copy over the network - and of course, all the empty space doesn't need to move).
Not really even touching on what private/hybrid cloud means to an enterprise; but suffice it to say, there's a reason why nearly every company in the Fortune 500 has some users pulling out corporate cards and buying compute from AWS; and why they'd like to supply that same experience to their users on a private cloud platform.
Lots of people write books to cash in on the latest marketing trend. There's been many names for this type of computing. These days everybody attaches "cloud" to everything they do, but there's nothing original here.