Defining an SLA
10 · JUN ·2011
In the business world, network redundancy has been high on the agenda post 9/11. We’re in an age where businesses are increasingly reliant upon technology to underpin their day to day operation; prolonged network failures are more costly than ever before.
One common indicator of a connection’s reliability and resilience is the SLA, and in particular - the uptime guarantee. SLA’s are a useful guide to how robust a service is, they provide a reflection of a company’s confidence in the service they deliver and should provide assurance that the connection supplied is of a certain grade, and augmented by fixed response and fix times and round the clock support.
However an SLA’s uptime promises can sometimes blind organisations into assuming that they’ve mitigated against all points of failure and failing to look into their network topology. Consider this; if businesses were to quantify the cost of having business critical systems suffer an outage, are these costs compensated by the SLA on offer?
Five years ago, Fluidata brought an innovative aggregation platform to market called PureFluid. This bonds multiple circuits and/or technologies (ADSL/ADSL2+/SDSL/ 3G) from multiple providers, all lines deliver over the same IP subnets allowing for seamless failover in the event of a fault on one of the circuits or networks.
For this solution Fluidata provide an SLA of 99.99% in terms of service uptime and network availability. We believe, that perhaps with the exception of diverse fibre digs, we’d be hard pushed to provide a more resilient service and as such a higher SLA. After all no connectivity service is infallible.
Yet 100% SLA’s do exist in the industry, providing promises that are hard to keep. Typically when you investigate the design of these solutions, the points of failure and Terms and Conditions’ that back them up are not sufficient. All I am saying is beware the semantics and drill down to what’s actually deliverable and why.