Windows Server 2012 Supports Data Center TCP (DCTCP)

In the grand effort to make Windows Server 2012 scale above and beyond the call of duty Microsoft has been addressing (potential) bottle necks all over the stack. CPU, NUMA, Memory, storage and networking.

Data Center TCP (DCTCP) is one of the many improvements by which Microsoft aims to deliver a lot better network throughput with affordable switches. Switches that can mange large amounts of network traffic tend to have large buffers and those push up the prices a lot. The idea here is that a large buffer creates the ability to deal with burst and prevents congestions. Call it over provisioning if you want.  While this helps it is far from ideal. Let’s say it a blunt instrument.

To mitigate this issue Windows Server 2012 is now capable dealing with network congestion in  a more intelligent way. It does so by reacting to the degree & not merely the presence of congestion using DCTCP. The goals are:

  • Achieve low latency, high burst tolerance, and high throughput, with small buffer switches (read cheaper).
  • Requires Explicit Congestion Notification (ECN, RFC 3168) capable switches. This should be no showstopper you’d think as it’s probably pretty common on most data center / rack switches but that doesn’t seem to be the case for the real cheap ones where this would shine … Sad smile
  • Algorithm enables when it makes sense to do so (low round trip times, i.e. it will be used inside the data center where it makes sense, not over a world wide WAN or internet). 

To see if it is applied run Get-NetTcpConnection:

image

As you can see this is applied here on a DELL PC8024F switch for the CSV and LM networks. The internet connected NIC (connection of the RDP session) shows:

image

Yup, it’s East-West traffic only, not North-South where it makes no sense.

When I was prepping a slide deck for a presentation on what this is, does and means I compared it to the green wave traffic light control. The space between consecutive traffic lights is the buffer and the red light are stops the traffic has to deal with due congestion. This leaves room for a lot of improvement and the way to achieve this is traffic control that intelligently manages the incoming flow so that at every hop there is a green light and the buffer isn’t saturated.

image

Windows Server 2012 in combination with Explicit Congestion Notification (ECN) provides the intelligent traffic control to realize the green wave.

image

The result is very smooth low latency traffic with high burst tolerance and high throughput with cheaper small buffer switches. To see the difference look at the picture   below (from Microsoft BUILD)of what this achieves. Pretty impressive. Here’s a paper by Microsoft Research on the subject

image

Experts2Experts Virtualization Conference – Vienna May 25th-27th 2012

I’m attending and speaking at one the of the best  small scale virtualization conferences out there. I’m talking about the Experts2Experts Virtualization Conference (E2EVC) organized by Alex Juschin for many years now. I’ll be speaking at the conference on “Making Sense of  RSS, DMVQ, SR-IOV, RDMA and other advanced networking features”. We’ll see where Windows Server 2012 & the new generation of Hyper-V is at in regards to these technologies, how it stacks up against some other solutions and what looks promising. In other words what we are looking at to use in real live once Windows Server 2012 goes RTM.

I have the good fortune to attend some pretty big, impressive & high quality industry events. These are excellent places for networking and getting up to speed with the latest of the greatest form the big vendors and the ecosystem around it. But they are pretty expensive and large scale, so most people are so crazy busy at those you often miss out on some of the interaction, there is just to much going on.

E2EVC is special and adds a different kind if value that goes beyond its low cost. For one, nobody is trying to sell you anything. All attendees and all speakers are IT Pro’s that design, build, work with and support the technologies that are discussed. Hence the name, Expert 2 Expert. It’s a reality check on what are people really using, trying, evaluating. You’ll see what is really hurting us and what really works.  An event like this isn’t driven by marketing. It’s driven by interests, passion for technology and even more important from a business perspective the solutions they can and do deliver in real live. This proves that you don’t need to charge premium prices to keep the riff raff out. The fact that 2 days of this conference are in a weekend tells you the attendees are going there with intend and purpose.

The guys & gals attending & presenting are top notch. They don’t look  like slick advisers and analysts. It’s all very informal and relaxed. But make no mistake, these people are sharp and at the top of their game. Discussion and interaction is stimulated and lively. The aim is not to breed or create rock star speakers but to get people to share their experiences and knowledge. And here in lies the value. I really commend Alex Juschin for having succeeded in this.