I think ICANN (and any escrow organisation named by ICANN) should have a zone copy of the 2nd level TLDs (root being the first level). This would ensure continuity of the DNS in case of failures. The business continuity is the responsability of the registry and/or registrar. As a side benefit it would allow ICANN to publish stats on the domain names and check that zones are correct (technically that is). Cheers On 4/24/07, info@dnsportugal.com <info@dnsportugal.com> wrote:
Kind of tricky business this...
Looks clear that registry failure is something like an "Extition Level Event", so each and every step should be taken to avoid it from happening altogether. But as any "Act of God" there is no sure way to prevent that it could happen some day.
We should start deploying failsafe structures behind today ( or should i say ten years ago? ).
One suggestion appearing on ICANN blog comments is the creation of a "Backup registry operator". I do support the idea but it has some problems that need to be clarified.
One thing is the decision on what model should registry operate. Should it be a thick or thin registry. The thick model is robust when catastrophic things happens at regsitrar level ( take the Registry Fly debacle as an example ). On the other hand the thin model appears to be more robust when dealing with registry failure as data is distributed through so many places. I do not have a definitive opininon of what model should we choose, since both have goods and bads.
Maybe some kind of hybrid? One thing is clear though and that is the need for data redundancy, something that could be attained by pu in place strong data escrow both at regsitry and registrar level
Rui Bebiano
On 23 Apr 2007 at 14:53, RJGlass | America@Large wrote:
When Hurricane Katrina hit New Orleans, it knocked out a registrar,
DirectNIC. I happened to
have some stuff hosted with them and no longer do. Apparently it knocked out their DNS servers.
If a registry should fail, it would cause a much more dramatic failure of the system. There definitely should be a plan in place for such failures, including remote backups, maximum downtime, and a plan to replace the organization handling it.
-Randy Glass A@L
On 4/23/07, Jacqueline A. Morris <jam@jacquelinemorris.com> wrote: Dear All ICANN's Registry Failover project has recently provided an update to the ICANN Board and also gave a presentation to the ccNSO in Lisbon. The GAC Communique released in Lisbon also contains language on registry failover. In an effort to reach out directly with ALAC and users, they have posted on the ICANN blog some of the types of questions that the community, users and public may want answered in the event of a registry failure. See http://blog.icann.org/?p=105 . Please comment.
Thanks,
Jacqueline
No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.463 / Virus Database: 269.5.9/773 - Release Date: 4/22/2007 8:18 PM
_______________________________________________ ALAC mailing list ALAC@atlarge-lists.icann.org
http://atlarge-lists.icann.org/mailman/listinfo/alac_atlarge-lists.icann.org
www.alac.icann.org www.icannalac.org
-- ------------------------- AmericaAtLarge.org RJPacific.com DDMF.org
_______________________________________________ ALAC mailing list ALAC@atlarge-lists.icann.org
http://atlarge-lists.icann.org/mailman/listinfo/alac_atlarge-lists.icann.org
www.alac.icann.org www.icannalac.org
-- -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Franck Martin franck.martin@gmail.com "Toute connaissance est une réponse à une question" G. Bachelard