Fun little outage here last night. It was only something like 5 minutes without Internet service, but when it came back, DNS took about 30 minutes to start working. My router (the one I control, not the modem/router the ISP provides) currently uses OpenDNS and Google DNS *, neither of which was passing through the ISP router at first.

Eventually, queries started passing through, but at first, it needed to be preceded with a

`dig @8.8.8.8 example.com` (or `nslookup example.com 8.8.8.8` on Windows machines) followed by `dig @192.168.1.1 example.com` ( `nslookup example.com 192.168.1.1` on Windows) to get desired sites back into the local DNS caches.

It was totally weird, and yes, DowntimeDetector and similar sites had big red smudges over SoCal and part of AZ, but the ISP’s own site said no outage was known.

* There is a new owner, but the prior owner of the local provider used to force DNS queries to randomly fail, so that it could redirect customers to a Yahoo search page ... presumably Y! was paying a fee for each link clicked on said page. Among the queries that frequently failed were Twitter and Facebook—and Google’s search page—so it was obvious they were doing it intentionally to jack up their revenue.