A recursive DNS query is a request from a client for a website that must be responded to with either the sought response -- the IP address associated with the corresponding site name or uniform resource locator (URL) -- or an error message stating that the site does not exist.
Recursive DNS queries are standard requests by users or by domain name system (DNS) servers that are configured to pass along unresolved IP requests to another DNS server; in that way, the first server can stay available. Recursive DNS queries and recursive DNS servers are used to distribute some of the huge workload involved in resolving the names of websites into URLs.
Recursive DNS requests are filled either from the server’s own zone file cache or by consulting other servers in the hierarchy of authoritative DNS servers. The client request does not get passed on, however. It is the recursive DNS server that stays in contact until it provides the IP address or finds the request cannot be resolved and returns an error message stating that the site does not exist.
It should be noted that all DNS servers must support iterative requests. Even the requests made by recursive servers to other DNS servers are iterative. At the level of root DNS, all servers are iterative as they are the end of the line and a crucial point that needs to be available for the internet to function.