- The workshop's program and speakers have been published.
- Register to participate in the workshop via live stream. Recordings will be available for those not able to participate in real time.
- The keynote speaker will be Bruce Schneier. Schneier will discuss the risks and mitigations of domain name collisions.
The Internet research and engineering community has long been aware that installed systems in the Internet query the DNS root for a wide range of top-level domains (TLDs) that are not delegated and that therefore result in an error return (commonly called NXDOMAIN) for the query. NXDOMAIN queries are common for non-root domains as well. In the ICANN community, it was observed as early as the Security and Stability Advisory Committee (SSAC) report SAC045 that the addition of new top-level domains in the DNS root could result in name collisions for installed systems. Some installed systems, following established operational practices, generate queries to the global DNS with name suffixes that under reasonable assumptions at the time the systems were designed or configured, were not expected to be delegated as TLDs. The installed systems in many cases depend explicitly or implicitly on the indication from the global DNS that the domain name suffix does not exist. Once the gTLD is delegated, then the global DNS may give a different response to a query involving the gTLD than it did prior to delegation. The impact of the global change on installed systems will be varied; risks to installed systems introduced by name collisions may arise due to varied causes.
Until recently, there was only a small amount of research into name collisions with new delegations, in part because the addition of new delegations was relatively infrequent. In 2013, this has changed. The ICANN evaluation of over 1400 applied for gTLDS reached a milestone and new delegations into the DNS root have begun with a high frequency. Several studies in 2013 identified and quantified installed systems with high potential risks. Monitoring and mitigation approaches have begun to be developed and discussed.
There is broad field of open research questions about name collisions and about the behavior of DNS name resolution when there are multiple scopes or resolution contexts being used. Because the types and numbers of potential collisions are vast, and because today only a relatively small amount of rigorous investigation has been conducted, there is a pressing need for more research. This workshop will gather investigators from multiple arenas and it aims to advance the state of knowledge and practice around name collisions.