Their news section says ns1 only is down but ns2-4 appear to be down as well. I notice afraid.org actually uses pickle.afraid.org and evergreen.afraid.org and granite.afraid.org for their dns and they seem to be working fine.
Every 2 years or so it seems afraid.org has downtime issues. You can almost set your watch by it.
While I was aware and watching this event unfold before my eyes, flooded with emails and phone calls, I moved the IPs right from the beginning for my dns project (ns1 is still getting hit, even now) - one problem is propagation delays to DNS servers can take any amount of time ranging between hours, and days (depending on the ages of the caches that are closest to the client).
The issue was that someone brought in a heavily attacked domain last week while ns1 was experiencing a issue of its own a day earlier. That traffic disrupted all of the remaining servers (ns2 ns3 ns4) with traffic many times greater then the network cards were capable of during a time NS1 was coincidently under maintenance, so the use of the full capacity allocated was also coincidently not available during the network event.
If ns1. was not already under maintenance then its likely things would have been fine, but I'm not making excuses for what happened. Whatever the reasons, it happened - and I am quite upset about it. Years of trust building efforts can be destroyed in hours in an instant.
I have a solution I have been working on for this each day since then. The fruits of this effort will be available very soon. I invite you to judge its performance after the change!