Before the Browser Agent can begin to download objects or follow a transaction, it must first perform a DNS lookup to find the IP address of the website to be tested.
In general, end users would use the DNS servers of their Internet Service Providers to look up an IP address. However, the Synthetic Classic nodes do not work in this manner; they each contain their own dedicated server (running BIND) that handles any DNS queries. The IP address of the node is also the IP address of the local DNS server for that node.
Each node uses a caching name server to execute the DNS lookup. This DNS server interacts with the root name servers — the top-level domain name (TLD) — which in turn redirects these requests to the authoritative DNS servers. The authoritative DNS servers send back the IP information along with the TTL stamp (the Time To Live stamp, or the amount of time for which the IP will be considered valid).
Note that we cannot flush the DNS cache of any public node; our nodes obey the TTLs dictated by the DNS record of the domain in all circumstances.
For more information, see Backbone Nodes in the documentation.