Live data from Hacker News

Metal as a Service

maas.ubuntu.com

91–100 of 106 posts

Re: Metal as a Service

#91
post #55
post #49

Earlier quoted context omitted.

If you go look at what companies are running in the cloud, and what companies are hosting their clouds on, Canonical and Ubuntu are the 800lb gorilla in the cloud market. If you were talking naively about where Canonical should focus to make money you'd have to favour cloud tech over the desktop. It's not a simple as that though, those two facts aren't separate - Ubuntu got to be the platform of choice for the cloud…

I"m sorry but you are mistaken... if you look at server stats, majority of "the cloud" is RHEL/CentOS based. I'm not going to argue that there are not many people using Ubuntu for servers... but it's not the standard when you get to big deployments. (I think you are refering to Ubuntu being the default Linux choice on EC2 which certaintly helped it's popularity. However, EC2 is not even built from Ubuntu... it's RHEL…

I won't make it an argument, but I'll share what I know and see if you see things differently. You probably won't, but I'm inclined to say it anyway :-)

First, to clear some things up, I'll avoid conflating traditional enterprise servers and hosted servers with the on-demand, pay-as-you-use cloud model. The later is what I mean by "Cloud", not just hosted servers.

To my knowledge Red Hat only holds a lead in server OS installations with paid support, and that's almost all in the former model (non-cloud, for me). Incidentally that's also the stat you'll see when companies publish reports about Linux's market share in general - they often just look at who's buying Linux. It's hard to get a real picture of what people actually use when you include unpaid Linux usage (which is most Linux usage), but the best data I know of is created by scanning web hosts, and suggest that Linux server run in order Debian, Ubuntu, CentOS and RedHat:

http://w3techs.com/technologies/details/os-linux/all/all

Red Hat's sudden obsession with OpenStack in the last 12 months, the move to take over CentOS and the purchase of InkTank are reflections that Red Hat are struggling to get in that game and are prepared to spend big to do so.

Why are they struggling so hard? Because they're frightened. The steady, "we want everything to stay the same for 10 years" model they've been serving so well is seriously threatened and they need to be in the new market to continue their success. Businesses love the idea of the cloud (if not yet the reality). Read what companies say about the cloud - they'll talk about it in terms of flexibility, and opportunity, but what they really love is that it means that you can offload the genuinely generic work of IT infrastructure management. This reduces the reliance on staff and consultancy costs, which are the most expensive aspect of corporate IT (hint, want to stay employed, get with the program, fast!). It also means that whether you're using Red Hat or Ubuntu, you get to negotiate the fee for OS level support via a behemoth like HP, Amazon, Microsoft or Google, or if you're a huge company with a private cloud, you centralise all that and do it in bulk. The same goes for the physical space for hosting.

Now Amazon were an existing Red Hat customer and so their servers were already running that when they started on this public cloud adventure - which afterall was a clever way to reuse their existing investment in infrastructure and personnel outside peak load time. That's not a big win for Red Hat, but there's growth of their existing business for sure (who knows what kind of deal Amazon got from them?). Still, the utilisation of that cloud is the bigger chunk of new business there, and it's Ubuntu winning the majority of it.

Now spread out to the nascent cloud market as a whole, sure Amazon are a huge public cloud with the bulk of utilisation, but there are many others, new ones all the time. Additionally and many, many private clouds that are being created those are the opportunities for new revenue in the hosting market. Almost all of them are going with OpenStack and 55% of OpenStack deployments are Ubuntu right now (I've seen Red Hat running at less than 1/10th of that business. I'd expect Red Hat to increase that number a lot in the next couple of years (they'll convert some existing customers for sure), but don't be fooled by the recent, aggressive press releases from Red Hat - they're fighting to get a foothold in this market, not controlling it by any means.

Re: Metal as a Service

#93
post #40

Earlier quoted context omitted.

Fair enough. I was totally convinced that Amazon had baremetal instances available, but now I cannot find anything, so that could probably be a figment of my imagination.

All of Amazon's EC2 compute instances are run on top of Xen; there is an interesting Reddit AMA post with a lot of detail: http://www.reddit.com/r/IAmA/comments/1e5o4p/iaman_exaws_eng...

That is a very interesting thread, thanks. I think I was confusing dedicated tenancy with baremetal. Dedicated tenancy of course still runs on Xen.

Re: Metal as a Service

#94
post #89

We spent around 6 weeks trying to wrangle Juju and MaaS into a working state in August 2013. Skimming through my notes we surmised the following. * Auto-enrollment of nodes was tough to get working. * Overlay and config management was weak and better handled by Puppet. Though Juju beans were touted as able to handle this. * Juju 1.13 and maas 1.3 do not support isolated juju environments in the same maas cluster. No…

On a somewhat related note, here's a couple of links I found when looking around for more info on maas and juju: Virtme: a very promising start to get a proof of concept up and running -- sadly it appears to be abandoned?: https://manage.jujucharms.com/~virtual-maasers/precise/virtm... http://javacruft.wordpress.com/2013/06/25/virtme/ A (partial) list of manual steps that try to achieve the same (with the goal of end…

As the author of the 2nd link there, I'd advise some caution, it's old and not complete, I will update it at some point soon when I have time to run through my whole process again.

Re: Metal as a Service

#95
post #89

We spent around 6 weeks trying to wrangle Juju and MaaS into a working state in August 2013. Skimming through my notes we surmised the following. * Auto-enrollment of nodes was tough to get working. * Overlay and config management was weak and better handled by Puppet. Though Juju beans were touted as able to handle this. * Juju 1.13 and maas 1.3 do not support isolated juju environments in the same maas cluster. No…

On a somewhat related note, here's a couple of links I found when looking around for more info on maas and juju: Virtme: a very promising start to get a proof of concept up and running -- sadly it appears to be abandoned?: https://manage.jujucharms.com/~virtual-maasers/precise/virtm... http://javacruft.wordpress.com/2013/06/25/virtme/ A (partial) list of manual steps that try to achieve the same (with the goal of end…

You should probably start with Juju docs on setting up an environment with MAAS as a "provider"

https://juju.ubuntu.com/docs/config-maas.html

Re: Metal as a Service

#96
post #4

This is an interesting concept and a good addition to all the IaaS offerings around. Just one year ago, while doing my Master's thesis, I pondered the possiblity of creating an IaaS that didn't involve virtual machines but physical servers instead. This should be great especially for high performance and special hardware needs. Next step would be all sorts of mixed blends between virtual and physical servers: real CP…

You've reinvented what we used to call dedicated servers. There is nothing new under the sun. Might as well call it physicalization. :)

:-) Of course it's not inventing anything, it's just providing a convenient way to commission those servers that can also be driven by other tools. If you bring a lot of server resource on line this is very handy. Juju, for example, can put things on MAAS servers, and one of the things it could put on there is a virtualised environment that could also be managed by MAAS and juju to deploy other workloads. MAAS knows some stuff about the hardware too, for example, how many network cards the machine has and on which networks, which means you can interrogate MAAS to find machines suitable for a given workload.

Re: Metal as a Service

#97
post #85
post #72

Earlier quoted context omitted.

I agree Canonical should stick to their core, but it's not the desktop. It's cloud. They've demonstrated they do not have what it takes to make desktop net profitable for them. They make most of their money by selling support for their cloud stuff — this is what they should focus on. Desktop has done nothing but lose them money.

Well, Canonical could always make their own laptops with perfect driver support and charge a premium for that. I don't know why they're not. As of right now the only laptops that consistently work well with Linux/Ubuntu are Thinkpads and Thinkpads suck for e.g. gaming. No one except Microsoft can make money selling an OS - and they can only do it because they have a massive advantage in user training and software eco…

Note that as of late both Apple and Microsoft are making moves towards "the OS is free" for certain markets. I hear tell that Microsoft are even paying OEMs to put Windows on phones.

Re: Metal as a Service

#98

I was disappointed to learn I can't RESTfully request the Sepultura-like cover of George Michael's "Careless whisper" that keeps playing in my head.

Jono Bacon was responsible for that kind of Metal As A Service in Canonical - he's moved on to the X-Prize now, you should try there ;-)

Re: Metal as a Service

#99

Maybe someone with more familiarity can tell me how this differs from/expands on Cobbler?

Robbie Williamson, the man at the head of the responsible division of Canonical has documented how Cobbler and Puppet were used in the early days and how experiences there lead to MaaS being developed.

https://ubuhulk.wordpress.com/2014/03/10/canonicals-office-o...

Re: Metal as a Service

#100
post #79
post #70

Earlier quoted context omitted.

Wow, no. There's a ton of people that could use MaaS. People use the big public clouds because they're easy. With MaaS you can build your own cloud almost as easily, and skip the Amazon tax. It's great for big companies that want total control, it's great for any company that doesn't want their sensitive data in someone else's server room (there are a lot of these in finance and enterprise), and small companies that…

You're right, but honestly a lot of those enterprise rationalizations about their "sensitive" data are just defending headcounts and turf. Google and Amazon probably do a better job securing data than 90% of enterprise sysadmins.

I know that, and you know that, but do they? In my experience, no. And also, does the SEC know that? Possibly not. There's a lot of laws about this stuff where if you screw it up, people can go to actual prison.
Post reply on HN