Evaluation of scalable, on-demand DNS-as-a-Service

The Domain Name Service (DNS) is a vital service in the Internet. Much more than a simple translation mechanism, it also allows higher profile functionalities such as load balancing and enhanced content distribution. In the scope of cloud computing, DNS is foreseen as an elastic and robust service, supporting failover mechanisms, decentralised configuration and multi-tenant isolation. This paper presents and validates a cloud-based architecture for DNS as Service, considering the expected principles of security, scalability and elasticity. The obtained results reveal that the proposed architecture is capable of accommodating a high-load of DNS queries per second by dynamically managing the amount of used resources, enforcing a constraint of reduced query latency (<; 1s). Additionally, it also incorporates failover monitoring mechanisms to ensure the stability of the system. The performed assessment in Fed4FIRE testbeds shows that this approach allows for reduction of operational costs, managing used resources according to the service's load introduced by the simultaneous clients and by appropriately scaling in or out DNS servers independently of the underlying cloud infrastructure platform (e.g. OpenStack, Amazon Web Services).