|
Reliability is attained by a combination of network design, operating standards, and measurement. Radiant measures its actual performance weekly and makes this information available to all interested clients.
Redundancy
The first path to reliability is to avoid single points of failure. All of Radiant’s core network elements are either internally redundant or are operating in High Availability (HA) mode (i.e. dual interconnected devices). This redundant design applies to the entire core technology stack:
:: M10i BGP core router (internally redundant)
:: SA6000 SSL VPN appliance (HA)
:: ISG2000/SSG550 Integrated Security Gateway (HA)
:: ERX 1440 Provider Edge router (internally redundant)
Radiant’s private backbone is fully redundant. So even if one connection fails completely we have enough capacity to move all intra-Radiant traffic on the remaining connections. And, our Internet connections are quadruply redundant from each data centre.
In addition, wherever possible, Radiant utilizes redundant aggregation circuits with our last mile providers. Unfortunately, this is not always possible, and last mile circuits are always subject to failures such as cable cuts and Central Office outages.
Failover Options
To sidestep these inevitable problems with last mile circuits, Radiant has developed a rich set of failover options. We can provide our clients with failover options ranging from dialup to very high speed fixed wireless. All of these options can be made seamless – the failed location may not even realize that there is a problem – and carrier diverse, so that a local CO problem can be worked around.
|
