DNS Failover: One Trip Shouldn’t Take You Down
Downtime for revenue generating sites almost always means lost income. E-Commerce stores, especially those with international clients, depend on constant uptime. Sites that rely on advertising revenue aren’t going to be generating ad clicks and impressions if users can’t access the content. For essential business operation services, huge costs can be incurred in man-hours and lost time if servers aren’t accessible.
It’s been estimated that downtime can cost the average business $138,888 per year, but many of the costs of downtime are not so obvious as lost sales or advertising revenue. Downtime for web services can sometimes necessitate payouts to disgruntled clients who pay to be able to access those services when they need to.
Downtime can also cause considerable damage to a company’s reputation, especially given the prevalence of social media. We’ve all seen less than temperate customers berating businesses for even minor service interruptions; often a golden opportunity for competitors.
Infrastructure with a single point of failure is intolerable in the modern enterprise environment, which is why most business strive to provide multiply redundant systems. Hardware will fail, staff will make mistakes. It’s not a matter of if something will go wrong, but when it will go wrong and what procedures are in place to mitigate the effects of that failure.
DNS is the first point of contact between web clients and the sites and services they connect to. As such, it’s an effective point in the network at which to implement failover systems.
DNS failover is simple to understand. A DNS hosting company implements monitoring nodes which periodically check the responsiveness of servers. Monitoring is usually carried out by multiple nodes in diverse geographical locations to ensure that network conditions local to one monitoring node do not trigger the failover process. If more than one of the nodes detect that a server has become non-responsive, then new DNS records can be propagated throughout the system, using technology like DNSMadeEasy’s Peregrine Instant DNS Propagation. The unresponsive server is removed from the pool of servers that users are directed to, and the IP of the faulty server is replaced by that of a functioning server, which is also monitored in the same way.
Because monitoring is continuous, as soon as the problem with the primary server is resolved and it becomes available again for serving content, the DNS failover system can switch back to that IP.
With quick DNS propagation and appropriate TTL settings, using DNS failover procedures can provide an effective first line of defence against downtime incurring infrastructure failures.