Microsoft Windows Server 2016 Moving To Per-Core Licensing (arstechnica.com)

rbrandis writes: Windows Server 2012 has two main editions, Standard and Datacenter. They had identical features, and differed only in terms of the number of virtual operating system instances they supported. The licenses for both editions were sold in two-socket units; one license was needed for each pair of sockets a system contained.

Windows Server 2016 reinstates the functional differences between Standard and Datacenter editions. Datacenter will include additional storage replication capabilities, a new network stack with richer virtualization options, and shielded virtual machines that protect the content of a virtual machine from the administrator of the host operating system. These features won't be found in the Standard edition.

Windows Server 2016 licensing moves to a per core model. Instead of 2012's two socket license pack, 2016 will use a 2-core pack, with the license cost of each 2016 pack being 1/8th the price of the corresponding 2 socket pack for 2012. Each system running Windows Server 2016 must have a minimum of 8 cores (4 packs) per processor, and a minimum of 16 cores (8 packs) per system.

Comments

It's almost like a fetish

[ZorinLynx]

Microsoft seems to have a fetish for making licensing complicated.

I suppose since they practically invented the concept it makes sense. But damn, how far can it go?

Re:It's almost like a fetish

[BarbaraHudson]

Microsoft seems to have a fetish for making licensing complicated.

I suppose since they practically invented the concept it makes sense. But damn, how far can it go?

They didn't invent the concept. They're just following in the footsteps of Oracle, IBM, etc.

Re:It's almost like a fetish

[smittyoneeach]

Jeff Bezos is all: "Hey. You. Get off of my cloud."

Re:It's almost like a fetish

[mlts]

Here is the ironic thing. IBM with POWER7 has two modes for their chips. One is the usual functionality, where a 32 core CPU uses all 32 cores. The second mode, called TurboCore, disables half the cores... but allows the cores that are working to use the cache of their disabled neighbors, as well as run the CPU at a higher clock rate.

The reason for this mode is because Oracle, Sybase, et. al., all have per core licensing for production systems. So, having the ability to turn off a good amount of cores will cut the fee in half, and that licensing fee can be very substantial.

One advantage of Microsoft was that they licensed per CPU socket. Now, in Windows Server 2016, that changes... and I'm not surprised it did, just because of the amount of cores available on Xeons and AMD CPU chips.

Maybe this is a good thing. Customers will demand that Intel and AMD start having more oomph per core than just adding more cores to the die. This will help a lot in tasks that can't be multithreaded (fast fourier transforms if doing video, for example.) Maybe we will see the IBM TurboCore mode (not to be confused with AMD's TurboCore) used in the amd64 architecture.

Re:It's almost like a fetish

[Rockoon]

This will help a lot in tasks that can't be multithreaded (fast fourier transforms if doing video, for example.)

There have been parallel FFT's algorithms for years that scale fairly well, especially for multi-dimensional data (3D transforms get an almost a linear scaling with core count.) What the hell are you talking about?

Re:Cores? Packs? Sockets?

[Anonymous Coward]

They mean you should run your infrastructure and business-critical services on Linux or BSD so that Microsoft can't hold your entire company hostage at will.

Any real tangible merits to using Windows Server?

[bogaboga]

Can someone tell me what exactly I am missing by [stubbornly] refusing to use Windows Server? I know there surely exist some advantages but what are they really?

I have been using Debian Linux on our servers for almost 13 years now and we have no regrets! We have Samba installed as well.

I sincerely do not know what I am missing as our systems have not given us any trouble for a long time.

I must say we have some company contracted for support just in case. Who will bite?

You may wonder what then keeps me busy: Well, We experiment a lot and contribute to quashing Debian specific bugs from time to time.

Not surprising

[JustAnotherOldGuy]

As someone who is peripherally involved with MS data centers I can tell you that the whole Azure/cloud thing is booming like mad. It's insane.

They literally cannot build data centers fast enough so what they're doing is buying and/or leasing buildings, gutting them, rebuilding them and hardening them to keep up with demand. And they're still not keeping up, there's a huge pent up backlog of demand and capacity that is growing like crazy. They literally can't keep up with the need for secured server space that meets their requirements.

Re:Any real tangible merits to using Windows Serve

[Anonymous Coward]

I know there surely exist some advantages but what are they really?

Lack of SystemD isn't advantage enough?

Vendor lock-in is the only explanation

[Jeremi]

If you're a sane businessperson, you make sure your server software is easily portable to any OS, so that when a particular vendor tries to hike their licensing fees, you can just say "thanks, no thanks" and move your software to some other platform as necessary.

Or, if you're completely blinkered and naive, maybe you've decided to irrevocably tie yourself and your company to a single vendor's platform, so that they can now do whatever they want to you and your only choice is to either pay up or rewrite your software from scratch.

If you find yourself paying lots of money to run your software on an OS named for and designed around its GUI interface --- in order to run your software on a headless server in the cloud -- you might be in the latter category.

AMD "modules"

[tepples]

Cores: the number of compute cores available. I think this is pretty clear.

Not on AMD it isn't. The cores of its processors since Bulldozer are sort of a hybrid between actual cores and SMT thread states.

Re:Any real tangible merits to using Windows Serve

[Zero__Kelvin]

"Microsoft knows what corporations want and makes it dead easy to do things that scale from a small business to a huge multinational."

That is a ridiculous claim. I have extensive experience with Linux and Microsoft, and claiming that Microsoft makes things easier is just plain ridiculous. It is the kind of claim that could only be made by a person who has Microsoft experience, but none with Linux (or at least significantly less).

Re:Any real tangible merits to using Windows Serve

[mlts]

I think "making things easier" is being mixed up with "easier to find MS experience than Linux experience."

The problem I encounter, having sat in both worlds, is that each side thinks their stuff is the right hammer, and everything is a nail. The MS guys want to use their wrench as a screwdriver, while the Linux guys want to carve notches in bolts so they can use their screwdriver in place of a wrench.

A couple use cases: Spawning Hadoop instances on OpenStack [1] or AWS is a lot easier with Linux than Windows. It can be done with Windows, but it is a lot easier to find howto guides and such under Linux. Another case is popping up nginx web servers on compute nodes for static content behind a load balancer. That is pretty easy with ansible [2], lsync, and varnish. In Windows, it can be done, but it would require some fancy footwork with SCCM/SCOM/WIM.

On the opposite side, for a massive directory service (something spanning multiple geographical regions, with many employees and company division/org charts that look like spaghetti), AD has a lot more support than the various LDAP platforms [3], and has proven to be good enough, security-wise.

Best thing to do is use both. Windows winds up at the core, Linux/BSD/etc. are at the edge.

[1]: Windows and OpenStack are like oil and water. I've not heard of any OpenStack deployments based on Hyper-V, especially on Kilo and Liberty. I wouldn't be surprised to see it (as Microsoft has embraced Docker in a useful fashion), but not at this stage.

[2]: Ansible is easy to include in the VM image, so it either can have an image pushed to it, or it can hit a Git server, grab its playbooks, then run those.

[3]: I've used other directory services. I would say that AD is a lot less painful than AFS or DFS/DCE. Things can change on a dime, and an AD competitor that can scale and replicate can come out of nowhere, similar to how Ansible/Puppet/Chef/Salt wasn't on anyone's radar a few years ago, but now is a staple of IT/DevOps as of now.

Technical solutions to price fixing?

[unixisc]

More oomph per core also violates a major requirement of all post 2000 CPUs - that they conserve power. In the days before multicores, you had CPUs trying out various combinations of superscaling and superpipelining in order to maximize performance. A major reason being that OSs at the time had limited multi-processing capabilities, and even when they did, their software didn't.

Things changed once NT came around, and since NT could do SMP, Intel could boost performances by tossing a number of their top core CPUs into the mix, and NT, being SMP like Unix, could handle that. So now Intel had a new more scalable way to boost performance, as well as segment the market instead of sinking w/ the Itanic. They could offer dual or quad core for PCs, while offering their 8-32 cores for servers.

The GP's description was good, but the problem w/ that approach is that it's a technical solution to an artificial problem - that of hiking prices by changing the pricing model. Unlike technical solutions to issues such as power consumption or limited performance, this is not something that the technologists should be solving. The proper solution to Oracle, Sybase and the other enterprise software companies jacking up prices is to explore more FOSS solutions, such as ProgreSQL or NoSQL. And when Microsoft does this, explore the BSDs or Linux.