If you talk to any organisation in New Zealand today, each and every one of them is facing a similar set of business challenges. Corporate networks need to support more data and network traffic with less resources and budget, and they need to do this much more quickly than ever before in a heightened security environment. However here in New Zealand we have a smaller business footprint and can be more nimble and adopt new technologies that can help overcome these challenges.
We are seeing this trend with the broad adoption of virtualisation and cloud computing technologies in many organisations. Underpinning every conversation we have with partners, resellers and end users is the drive to optimise information technology and assets to achieve higher business efficiencies.
With this radical evolution of the data centre and how applications are hosted in a virtual environment, there is a need to re-architect the traditional network in order to meet these business demands. This organisational driver has led to the development of the Ethernet fabric.
Old concept, new application
The idea of fabric networking is not new and has been discussed for some time in relation to high performance computing systems. Fabric networking has been used as a concept for years in storage area networks. In a nutshell, fabric networking links storage, networking and parallel processing functions with high bandwidth connections, moving data from one node to another. The network topology when viewed from afar appears as a of a grid or a fabric, hence the terminology. Industry experts have acknowledged that the changing demands in corporate networks will not only see a continued discussion about fabric networking, but many organisations will incorporate more fabric networking further into their systems.
From a technical perspective, the network must evolve and it must be simpler to operate, more flexible, highly resilient and much more scalable. These requirements are best met with the scalable fabric architecture of Ethernet fabrics.
Ethernet fabrics provide higher levels of performance, utilisation, availability and operational simplicity. With Ethernet fabrics, the network is easier to set up, operate and scale. The network must also be more flexible, be highly resilient and much more virtual machine aware.
Ethernet fabric networking is able to meet these challenges courtesy of its five major defining characteristics.
Standard datacentre networks use three-tiered architecture: access, aggregation and core. Traffic has to move up and down a logical tree to flow between server racks, adding latency and creating congestion on Inter-Switch Links.
Ethernet fabrics flatten the traditional network by collapsing the aggregation and access layers into one layer, thereby increasing both efficiency and performance while reducing capital costs.
Ethernet fabric also eliminates the need for Spanning Tree Protocol (STP) in the network which further reduces costs and improves network resiliency because multiple network paths are available among all the devices in the fabric.
In summary, an Ethernet Fabric gives enterprises the opportunity to build a flat, high performance, multi-path, deterministic mesh network for the data centre.
Traditional Ethernet switches require configuration of each switch port. This includes setting network policies such as QoS, security and VLAN traffic, etc. When only physical servers connected to the network, this model was sufficient. But today, server virtualisation requires multiple virtual machines to be configured on each switch port. When a virtual machine migrates either for load balancing or routine maintenance, the port configuration has to move to a new network port or the migration fails. This requires manual configuration.
Ethernet fabrics have distributed intelligence which allows common configuration parameters to be shared by all switch ports in the fabric. In the case of virtual machine migration, the network policies for that virtual machine are known at every switch port so migration does not require any changes to network configuration.
In an Ethernet fabric, switches share configuration information, and they also know about each other. When a device connects to an edge port of the fabric, all switches know about that device. As the device sends traffic to other devices, the fabric can identify the shortest loop free path through the fabric and forward frames with the lowest possible latency. New traffic types such as virtual machine migration and storage traffic are latency sensitive. Fabrics ensure this traffic gets to its destination with minimal latency.
Classic Ethernet allows only one path between switches. Improvements such as link aggregation groups (LAG) allow several physical links to act as a single link. This is manually configured on every port in the LAG and is often inefficient limiting bandwidth. If a new switch is added for more connectivity, it becomes increasingly more complex to manually configure multiple LAG connections.
Ethernet fabrics overcome this. When a new switch connects to the fabric, no manual configuration is required for the inter-switch links. The switch joins the fabric and learns about all the other switches in the fabric and the devices connected to the fabric. No manual configuration of policies or special LAG configuration on specific ports is necessary.
If multiple inter-switch links are connected between two switches, a logical trunk automatically forms. Traffic is load balanced in hardware so that utilisation is near line rate on every link for high efficiency. Should a link in a trunk go off-line, traffic on the remaining links is not affected and incoming frames are automatically distributed on the remaining links without disruption to the devices sending them.
As mentioned, the traditional Ethernet network uses STP to define a loop-free path, forming a logical hierarchical switch tree. Even when multiple links are connected for scalability and availability, only one link or LAG can be active. This lowers utilisation. When a new link is added or removed, the entire network halts all traffic for tens of seconds to minutes while it configures a new loop-free tree. This is highly disruptive for storage traffic, virtual machine migration, and so on. In the case of storage traffic, traffic disruption could cause a server crash.
Ethernet fabrics do not use STP to remove loops. They use link state routing with equal-cost multipath routes, which always take the shortest path through the network. When a link is added or removed, traffic on other links continues to flow non-disruptively. Link resiliency is assured and full utilisation of all links between switches is automatic when the topology is changed without any manual configuration.
Classic Ethernet switches require management and this is something that we are all familiar with. Each switch has to be configured and each port has to be configured for protocols (STP, RSTP, MSTP, LAG, etc), VLANs, network policies, QoS and security. As more server racks are added, more switches are added at the top of rack, middle of row or end of row. Each requires configuration and none can share common configuration parameters.
Ethernet fabrics share configuration information among all switches in the fabric. When a new switch joins the fabric, it automatically receives common information about devices, network policies, security, and QoS. This simplifies network configuration, reduces mistakes and reduces operating cost.
Incorporating Ethernet Fabric into the Channel
Due to the changing landscape in data centre networking, there is a real opportunity for resellers to provide customers with significant ‘value add advice’ to help them evolve to a more efficient architecture. Whether it’s discussing the current storage environment with a customer to determine how they can integrate use Ethernet fabrics, or selecting the style of Ethernet fabric technology that you want to recommend to clients, it is important to understand the design principles behind the technology.
For resellers it is key to ask the important questions and not be afraid to keep asking the questions. Does the fabric networking technology integrate and also interoperate with the existing infrastructure? Do you believe the solution you favour is agile enough to meet key business imperatives? Will this investment demonstrate value to the organisation and its clients in turn?
You need to be clear in the options for Ethernet fabric networking. Do your homework and understand where the industry sits in relation to trends. Is the technology complementary to your aims and who else is following the trend? Another consideration is the level of support that the vendor can provide. Initiatives that allow the end user to install and evaluate the Ethernet fabric solution, and not merely make a buying decision based on a demonstration, are both desirable and available. Programmes that make implementation and enhancement projects financially attractive should also be readily at hand for resellers.
In conclusion, the primary point of interest for Ethernet fabric networking is that it is here to stay. Ethernet fabric technology enables organisations to transition to the virtualised data centre and cloud-optimised network structure and for many organisations this is the direction for the future.