[dns-wg] KSK Rollover Postponed

> On 28 Sep 2017, at 12:53, David Conrad <drc at virtualized.org> wrote:
> 
> As far as I am aware, nothing is on fire. Given the lack of time criticality, I would have thought it’d be more important to the technical communities to have more concrete data to present. Given propagation delays in non-technical circles, I believe ICANN comms felt doing a “normal” press announcement-style approach sooner rather than later and then providing more details to folks who were interested later was a better approach. YMMV.

David, I would have expected some sort of announcements via the usual suspects mailing lists would have been part of that “normal” communication approach. YMMV. Instead, I learnt about the postponement because of a hearsay comment about a tweet.

More details would of course be welcome. [And I’m sure forthcoming.] In due course. When the time is right. Blah, blah, blah. However if that info isn’t yet available for sharing, it shouldn’t have prevented announcements to the expected technical/operational mailing lists. I don’t understand why those lists weren’t told or how using those channels to get the word out could have been harmful.

> We made an announcement within a few hours of making the decision to postpone the KSK rollover and are proceeding to attempt to gather more information to inform the community. Do you feel the fact that we did not send that announcement to technical mailing lists despite not having that additional information have an operational impact on resolver operators?

Perhaps. There wouldn’t of course be an impact on resolving and validation. There may however be second-order effects. Responsible resolver operators may well have lined up a small army of staff to be on call and prepare for problems and/or TEOTWAWKI on the 11th. If so, they wouldn’t need to wait for that additional information so they could tell their helpdesks and ops people to stand down. It wouldn’t immediately matter to them why the rollover was being postponed, just that this was happening and that any resolver operator planning for that event could be postponed too. From that perspective I think it would have been better to communicate the news as widely as reasonably possible.