When you design and build your office network you appreciate that some things on the network will have a single point of failure. Your desktop computer or laptop for example, is likely connected to the network via an Ethernet cable and if that cable fails for any reason you will no longer be able to connect.
For devices within your control, within your office and that you can physically replace on your own, this single point of failure concern is minor. Keeping a supply of Ethernet cables handy is a trivial cost and of no burden to the business.
Now let's look at the broader network design, your office will have a switch and that switch will connect to your office server or servers. You may have a rack of server equipment somewhere in your building and that rack will have its own switch so that you can connect all the server devices into the network together. Then a connection is made between that switch and the switch that all the desktops and laptops (and WiFi access points) are connected to.
Something like this:
This is a good architecture as it allows you to control the server environment and office environment as separate security domains. So the office would have its own subnet, let's say 184.108.40.206/24 and the servers would be on a different one, let's say 220.127.116.11/24 forcing traffic between the two networks to route, allowing it to be controlled and monitored at the firewall.
In this design we can consider that the critical single point of failure from a network connection perspective is now the link between the the office switch and the server switch. This is an acceptable risk because the path taken by the cable is well known, and the cable can probably be easily replaced. Also it is unlikely to be damaged by some third party digging up the street or crashing into a roadside cabinet.
Now wind forward to modern IT infrastructure provisioning and consumption. Your business is leveraging the benefits of on-demand resources and pay-as-you-go IT services from a 'cloud provider'.
Have you considered where this mission critical cable goes in the new network design? It likely starts at the edge of the building and goes underground (or over the local power poles if you still have them on your street, popping up at a roadside cabinet like the one below and then back underground all the way to the local telephone exchange.
The point here is, this once very critical cable that the business accepted as rapidly repairable single point of failure, is now some 20 kilometers long traversing an unknown number of streets and interconnection points, susceptible to the next piece of ground work performed by the council or roading crew.
Is it still an accepted risk?
What is the 'return to operation' (RTO) time-frame in the event of a failure, be it spade fade or car versus cabinet, or upstream network failure. Most businesses are not aware that the default service level agreement (SLA) commitment on BitSteam2 (BS2) services is 48 business hours (4 business days based on the LFC working hours of 7am to 7pm). So a failure on Tuesday might see you back online by the end of the week.
How can your business mitigate this risk and reduce this rather untenable 4 day outage window down to something viable in the world of 'always connected' business services?
Having a secure and reliable path between your office and your servers gives you the peace of mind you need to guarantee that your staff and perform the functions that keep the business running. This connection represents a single point of failure and one that is largely outside of your control.
With vGRID SecurePath™ you get the benefits of a secure, reliable and well understood path between your office and your servers. The connection does not traverse the public internet space and you are not at the mercy of congestion causing major events such as video streaming the world cup or patch-Tuesday downloads for Windows.
vGRID SecurePath™ comes in multiple speeds and connection (media) types including cellular (high speed LTE). Click the link below and download the full product detail PDF today...