DNS – preparation for the future 11/6/2015 copyright 2002 Bill Manning Introduction to the DNS “There are no urgent DNS problems” – Hotz “Yet…” - Manning 11/6/2015 copyright 2002
Download ReportTranscript DNS – preparation for the future 11/6/2015 copyright 2002 Bill Manning Introduction to the DNS “There are no urgent DNS problems” – Hotz “Yet…” - Manning 11/6/2015 copyright 2002
DNS – preparation for the future 11/6/2015 copyright 2002 Bill Manning 1 Introduction to the DNS “There are no urgent DNS problems” – Hotz “Yet…” - Manning 11/6/2015 copyright 2002 Bill Manning 2 DNS lemas It’s a system. The whole is greater than the parts. Historical co-mingling of administration & publication no intrinsic value in a single delegation single servers may be unavailable Not so true anymore Broadest reach unfragmented IPv4 UDP = 13 servers 11/6/2015 copyright 2002 Bill Manning 3 DNS Resolution Today “” Query girigiri.gbrmpa.gov.au “” name server Refer to au NS Query girigiri.gbrmpa.gov.au Refer to gov.au NS name server Query girigiri.gbrmpa.gov.au Refer to gbrmpa.gov.au NS au name server au gov.au name server nz gov sg edu Query girigiri.gbrmpa.gov.au Query Reply Address of girigiri.gbrmpa.gov.au gbrmpa.gov.au name server sa ips gbrmpa resolver 11/6/2015 copyright 2002 Bill Manning 4 Protecting the integrity of the system think globally, act locally keep current adopt correct stratagies • • diversity is key – and is lacking Too many domains use too few servers. sparky.arl.mil & ns.eu.net follow RIR practice & use your neighbors? Need better reliability than “public servers” 11/6/2015 copyright 2002 Bill Manning 5 So, what have the root-ops done for you lately? RFC 2010 first cut ops doc. others refined triage on placement Y2K statement calls out guiding principles TSIG guidelines Test-bed for experiments 11/6/2015 copyright 2002 Bill Manning 6 That said… Yann listed some items of concern, lets look at them. Some technical issues can not be delt with in the current system house cleaning/upgrades are needed controlled experimentation? Non-technical issues may not have technical components 11/6/2015 copyright 2002 Bill Manning 7 4. Technical Issues 4.1 Volunteer-based Operation: Is current system appropriate for the future? 4.2 Do we need any service level agreement(SLA) to be implemented? 4.3 Is the number of the root name servers appropriate for the future? 4.4 Is the current arrangement robust enough? 4.5 Do we need any architectural improvement? I.E. Distributed root name servers 11/6/2015 copyright 2002 Bill Manning 8 5. Policy & Management 5.1 Funding: Is the current funding scheme based on the volunteers appropriate? 5.2 Contract: Can we have an appropriate contract based on the current volunteerbased operation? 5.3 Is current scheme appropriate politically and managerially? Geographical diversity of the root name servers Master copy location - relocatable? 11/6/2015 copyright 2002 Bill Manning 9 6.2 Serving nonASCII labels Do we want to keep the internationalized Top Level Domains in the same server as the ASCII Top Level Domains? 11/6/2015 copyright 2002 Bill Manning 10 So, what do you want from the root operators? reliable service? guidance on technology adoption? more involvement in local issues? joint learning experinces? 11/6/2015 copyright 2002 Bill Manning 11 What do we want from TLD operators? Shared vision: Make the system robust, resilient Be responsive to change Serve accurate data Add value to our users & admins Minimise MTR, compartmentalize outages 11/6/2015 copyright 2002 Bill Manning 12 Questions? 11/6/2015 copyright 2002 Bill Manning 13