(Sometimes, RFC 1918 resolvers actually forward to Google Public DNS or another public resolver.)
My point about it being general outage referred to the fact that many users could not get answer from route53, even if they directly contacted aws without involvement of Google. DNS during the outage: https://pulse.turbobytes.com/results/5adf25e4ecbe40692e003abb/ DNS now: https://pulse.turbobytes.com/results/5adf2a6becbe40692e003aee/ MTR during the outage: https://pulse.turbobytes.com/results/5adf2844ecbe40692e003ad2/ MTR now: https://pulse.turbobytes.com/results/5adf2a60ecbe40692e003aec/
205.251.192.0 205.251.193.0 205.251.195.0 205.251.197.0 205.251.199.0
instagram.com simply had the unfortunate misfortune of having all their assigned nameservers in the block Joseph mentioned, and perhaps Google using HE upstream. The domain I included in the results above was lucky to have 2 nameservers (205.251.194.143 and 205.251.196.84) not included in the above list. On Tue, Apr 24, 2018 at 8:04 PM Stephane Bortzmeyer via Outages < outages@outages.org> wrote:
On Tue, Apr 24, 2018 at 02:49:17PM +0200, Stephane Bortzmeyer <bortzmeyer@nic.fr> wrote a message of 39 lines which said:
Feels like a general outage, not specific to 8.8.8.8
I tend to disagree. See with 25 RIPE Atlas probes:
With the default resolver (sometimes Google Public DNS, sometimes not):
If you look at the resolvers which SERVFAIL:
% blaeu-resolve -r 25 --displayresolvers -4 instagram.com ... [ERROR: SERVFAIL] : 3 occurrences (resolvers ['10.0.0.194', '192.168.12.254', '8.8.4.4']) Test #12294305 done at 2018-04-24T12:50:42Z
(Sometimes, RFC 1918 resolvers actually forward to Google Public DNS or another public resolver.) _______________________________________________ Outages mailing list Outages@outages.org https://puck.nether.net/mailman/listinfo/outages