In round robin scheme, DNS resolves address of first cluster for first DNS query, of second cluster for second query and so on. After giving addresses of all servers, it starts resolving address of first server again.
Round robin selection policy is very popular DNS scheme. It is used to equally distribute load on multiple servers of same capacity if it is assumed that all the clients generate same number of requests. But in practice, many clients generate very high or very low load thus resulting in load skew.
Average response times as reported by webstone is plotted in Figure 5.1. It is quite clear that there is not much variation in average response time as servers were never bottleneck in service of requests and their service time did not change much.
