OPEN QUESTIONS
Status | Board Question #2 related Questions |
| from meeting #47 | What is the role of the addition of applications incorporating stub resolvers directly rather than depending on recursive/iterative resolvers |
Status | Board Question #3 related Questions |
signaling interruption (how will apps and/or program logic change | |
| rae there harms that can be prevented through contracted relationships (i.e., ICANN contracted parties)? | |
Are we giving people incentive to work around the concept of one authoritative root? |
Status | Data Related Questions |
| Is DITL data sufficient to analyze long term trends? | |
| What additional signal is gained by adding additional roots? (in other words, What is the profile of query data from other roots besides A & J?) | |
| Given a set of strings observed at the root from an open recursive resolver, how does the traffic volume and other characteristics appear at the recursive? (in other words, What is the profile of query data as seen from resolvers that query roots? i.e. popular resolvers such as Google, Quad9, OpenDNS?) | |
| What bias is there in the geographic location of query data? Do query profiles differ by region or ISP? Do each roots have a bias in the query traffic they see due to peering agreements or anycast instance placement over other roots? (minor question) | |
| Give a set of strings observed at the root that are mainly Qname minimized, how does the traffic differ at the recursive? | |
| What was the change in query profile traffic (before and after) when some of the newer TLDs were delegated? | |
| Given the set of ICANN collision reports, is there any signal pre/post-delegation of the string that in hindsight could have indicated name collision risk? | |
| How do new strings (Crypto, Eth, etc.), that have purposely elected to use non-delegated TLDs, occur in the public global DNS? | |
| Where do the low, medium, and high risk strings of 2012 appear in overall traffic volume distribution? | |
| What is the query profile from within an organization that uses one of the TLDs under analysis (e.g. .corp)? | |
Status | Data Analysis Related Questions |
| What are the P-values (significance) and the effect size of the factors identified in the PCA analysis? | |
| What is the effect of QNAME minimization on the analysis? | |
| Update the table of top strings measured by query volume, distinct SLDs, distinct networks, and ASNs to include all of top X strings. | |
| All of the data analysis conducted thus far has been largely query-based. Is this sufficient? |
Status | Study Methodology Related Questions |
| Can an experiment be conducted by one or more resolvers or servers to measure the effect of delegating a name under analysis (e.g. create a view in partnership with an ISP and it’s customers to delegate .mail, .internal, or .corp)? |
Status | Risk Analysis Related Questions |
| Given a set of strings and a set of DNS-based risk measurements, how does Monte Carl simulation of random data sources influence risk assessment? |
