MoinQ:

1. watchA

Contents

    1. watchA
  1. aws bug

Peter van Dijk peter.van.dijk at powerdns.com Thu Sep 30 18:00:33 UTC 2021

Judging from dnsviz, a DS was present in the .com zone for slack.com
around 15:25 UTC today, and records inside slack.com were correctly
signed with the related KSK/ZSK set.

Judging from the DS as I see it coming out of some resolvers, the DS is
about 15 hours old at this point (so, introduced around 03:15 UTC I
think). Those cached DSes still have 10 hours to go.

However, the DS is gone from the .com zone, and the DNSKEYs and RRSIGs
are also gone. Hence, slack.com is Bogus for a big part of the Internet
viewer population.

https://lists.dns-oarc.net/pipermail/dns-operations/2021-September/021346.html

Viktor Dukhovni ietf-dane at dukhovni.org
Thu Sep 30 19:19:24 UTC 2021

It suffices to cap DS RRSet TTLs well below the .COM TLDs 24 hour
default.  If they cap TTLs at say 1 hour, the issue would have been
resolved 1 hour after the DS RRSet was deleted from the .COM zone.

Of course the DS RRSet might also have been manually flushed from the
cache, as special treatment for "slack.com", given their popularity and
the verifiable deletion of the DS records.

A final possibility is automatic refresh of DS records for domains that
go bogus.

So there are a number of ways to reduce the impact duration to below the
full ~1 day maximum.

2. aws bug

https://lists.dns-oarc.net/pipermail/dns-operations/2021-September/021362.html

On Thu, 2021-09-30 at 15:31 -0700, Michael Sinatra wrote:
> Once the negative cache ttl expires (5 min according to 
> the SOA minimum)

It appears AWS DNS has a bug here - their negative responses advertise
the 900 second TTL on the SOA records in negative responses, instead of
the 300 second MINIMUM. This, of course, changes nothing about your
argument. (But it would be nice if AWS fixed this.)


CategoryDns CategoryWatch CategoryTemplate

MoinQ: watchNS/slack.com/dns-operations (last edited 2021-10-02 01:03:25 by ToshinoriMaeno)