TL:DR A recent project bought a low cost network for the data centre. It cost less one-third of the the market leader & half the cost of a well known merchant silicon vendors. As a result, it is planned to last for two, maybe three years before it will be replaced. From this project I learned that “fast & cheap networking” could make a big impact on new data centre designs and business attitudes. Plus it was much more satisfying as a professional project. I’m now wondering – is networking too expensive ?
Most people refer to the “Data Centre Network” as though it was a single network. In practice, data centres have a number of individual networks. Each network is specifically designed for a specific purpose and function. A typical data centre network design has about five individual networks that connect together to form the “data centre network” which many people fail to recognise. I’ll define these networks and then look at the future of data centre networks with overlays. What seems clear, today, is that networking will provide different networks for different use cases and the customer will decide.
When working with Server and VMware people, there is a fair amount of misunderstanding of what is happening in the network. The best technical explanation for what is happening in LAG is, of course described at Ivan Pepelnjak’s IPSpace Blog – vSphere Does Not Need LAG Bandaids – The Network Does while Chris Wahl talks about the server side for VMware but I wanted to add something to the debate.
Making diagrams aesthetically appealing with visual impact is better documentation. Choosing the correct fonts on network diagrams will improve your network diagrams significantly. Here is some work on how to choose a good font and some recommendations on the best free fonts for your machine.
Overlay networking has been around for a year or so now and the ideas behind it are well established. It was about 3/4 weeks ago while researching VTEP functionality in Dell and Arista switches that I realised I could build manually configured tunnels with VXLAN and get the same results as an EoMPLS x-connect with almost zero effort. More importantly, I don’t have to pay for expensive hardware that has MPLS functions or pay again for software licenses to upgrade with MPLS features.
Stumbled over “AgilePorts” feature in Arista products this week: Arista’s AgilePorts technology enables the combination of four 10GbE SFP+ interfaces into a single 40GbE interface leveraging the parallel lane technology present in the 40GBASE-CR4 and 40GBASE-SR4standards. With AgilePorts, each 10GbE interface emulates one of the four parallel lanes, which are then driven by a 40GbE […]
I’ve been working on a lot of diagrams lately and pondering how to represent network architectures. I’ve been reading The Visual Display of Quantitative Information to get some inspiration on different approaches. I continue to be fascinated by the power of a network diagram that is well thought out and visually pleasing. And this fascination has led to my own focus on different network diagrams. In this post I’m thinking out loud one the different ways to represent information.
I was commissioned by GigaOmPro to write a report on “SDN Challenges in Large Scale Deployments”. I spoke with a number of network and virtualization engineers about their perspectives on SDN, the challenges they faced and how they would use Software Defined Networking in their data centres. It was evident during the research phase is that many people are not clear on what Overlay Networking is and just how deeply Overlay Networking will change Data Centre architecture and especially the nature of the networking and security domains.
I’ve been reading a presentation from Sharkfest 2012 where a engineers from Microsoft are presenting on their
Microsoft’s Demon – Datacenter Scale Distributed Ethernet Monitoring Appliance. The whole presentation is interesting but this particular slide caught my attention:
I was going to call this article “Ethernet Switches for Virtualisation Engineers” but, really, everyone should have some understanding of the internals of an Ethernet switch. But particularly I want to focus on how multicast and broadcasts are handled in a high speed, low latency environment like a Data Centre Network.
It’s vital to understand that latency is critical to your application performance. It is common for a single transaction to take hundreds of round trips so a small increase in latency on each round trip has a large impact on the perceived performance. The client will send a chunk of data and wait for acknowledgement. Even setting up the TCP connection takes a few round trip – remember that TCP sessions are setup, and each data transfer is confirmed.
A modern network switch will have latency around 10 microseconds. The Cisco Nexus 7000 is about 8 microseconds & Brocade VDX 8770 claims less than 4 microseconds. There are many reasons why a switch can be faster or slower but I’ll look at a specific example
Remember, the latency interval is the time taken to receive a packet, decode the address, lookup the forwarding table, switch the packet (and copy it if needed) and transmit out of an Ethernet interface. That’s really fast processing. How does an Ethernet switch do this ?
Network Engineers have to manage a lot of information. Products, technologies, textbooks, study notes and research material as well as new protocols and features. Just simple tasks like keeping product manuals handy for 40 or 50 products is a real problem. How do you keep the information organised, referenced, accessible and useful ?
This three part screencast is about how I manage all the “inputs” so I don’t feel lost in information after many, many people asked.
Lots in insight for Enterprise people to consider when defending your network against DDoS attacks. At CloudFlare, an attack needs to get over about 5Gbps to set off alarms with our ops team. Even then, our automated network defenses usually stop attacks without the need of any manual intervention. When an attack gets up in […]
Saw this nifty documentation idea in the Cisco® Smart Business Architecture (SBA) guide today.
You can deploy some modern firewalls in Layer 2 mode such that they are transpart
In this post, I’m looking at network designs with ECMP cores using TRILL or SPB, I’m realising that STP is equally improved in terms of risk and performance by reducing the STP domain size which leads to better stability, reduced risk and impact mitigation
A lot of people have talked extensively about OpenFlow making significant changes to the networking business. In particular, many writers have focussed on the possibility that OpenFlow enables a choice of using low cost network equipment instead of the expensive networking equipment that we use today.
Well, that’s highly unlikely.
Just watched this video on the ASA-CX. Only one thing made my heart leap with joy. NO JAVA.
Cisco is known for shipping products early to deliver new features quickly. But this leads to a reputation for buggy code which has customers report bugs (and Cisco fixing them). This means that you should never buy a newly released Cisco product unless you are willing to take this risk. This post looks a my process for analysing this risk and then selecting an IOS version by performing a bug scrub. In this case, I’ve been asked whether the Cisco C3750-X switches are ready for live deployment.
Juniper QFabric is a new approach to Ethernet Switch Fabrics. When it was announced last year,it was noted that the underlying physical design is a completely different approach to building Switch Fabrics. Here I’m taking a loosely research based approach to understand how Juniper QFabric is different from all other approaches to the problem, and also a look at some of the challenges ahead.