• Is it really the end of the traditional Three Tier Network Switching model? YES it is.
  • Are we starting to flatten this topology as we move from 1 to 10Gbps network backbones? YES you are.
  • Can we eliminate a full switching layer, yet still implement speeds of 40Gbps+? YES you can.
  • But what are the planning considerations that should always be made from the physical infrastructure perspective? DON’T KNOW!  That tends to be the most common answer!

 

In recent years we are starting to see a genuine fast paced shift away from the old and trusted Core, Distribution and Edge switching environment.

For decades this traditional model has served us all well, but now there is a sense of real momentum to remove that distribution layer, providing a flat network topology.

So this new flattened network, also referred to as Spine Leaf creates a challenging high density core cabling environment. If installed correctly then ready for both todays and tomorrows network demands.

 

The spine (access) switch fabric works in a fully redundant failover mode, linking to the leaf switches at speeds of 40Gbps to form that resilient backbone.

Those core spine switches will then link via fibre connectivity to smaller but still mission critical leaf switches, with 10Gbps server connects and critical connectivity uplinks.

Building forward planning in to this already complex design can then ultimately provide switch capacity for rapidly increasing server to server traffic.

 

We are already seeing our clients invest huge amounts of money in to this proven approach, as the global networking giants continue to promote and introduce more new and exciting technologies.

However, what we are not seeing is the same level of investment (time and money) around the physical infrastructure to support these critical systems.

This subsequently has a direct affect with network issues, which in turn will impact the performance of the organisations business applications.

There is actually (well should be) a logical network, as well as a physical network infrastructure.  Without this you are going to have problems with any implementation.

 

Despite an abundance of best practice whitepapers, very often these lessons are over looked, resulting in a costly rectification further down the line.

Network design should always include the physical layer infrastructure (as well as the logical switching) as one big picture – from Day 1.

Surely everyone should need to make considerations around Power, Cooling, Space and Network Connectivity? 

Reality is that this is more often neglected and rarely meets the criteria of the required modern day Data Centre standards.

 

In our option this is often down to lack of knowledge, rather than the willingness to invest further in the initial project.

Most of the time an increase in CAPEX from the offset is easily obtained, after the initial installation a discussion regarding further OPEX funds tends to be far more difficult to absorb.

There is as much demand of the physical infrastructure as the logical.  It’s needs to be properly designed and correctly installed for a successful deployment and future transition.

You are then fully prepared for the future challenges of ongoing consolidation, virtualisation, convergence and cloud adoption.

 

So here at Patch Solutions we can certainly help with the physical infrastructure planning from Cisco Nexus and ACI equipment lists, amongst other manufacturers.

In partnership with Panduit we can design your floor plan, produce wiring diagrams, document clear physical and logical equipment layouts, for both immediate and future planning.

Most importantly, we can help ensure that you are ready and prepared, between us identify and gaps that might in turn delay deployment and work in partnership around the delivery, build and implementation.

Not just from day 1, but a long term partnership as your network grows and evolves.

 

Ultimately taking this proven approach will ensure that you manage the project risks from the offset, thus producing a high performance, scalable and reliable network.

Your new Data Centre infrastructure will be then be fully optimised, providing a flexible and scalable network architecture with much improved long term operational efficiency.

Maximising uptime, whilst at the same time drastically reducing the ongoing operational costs.

Russell Meehan

Operations Director, Patchsave Solutions Limited