[WEB4LIB] Domain name "propagation"

[Nick Arnett]

At 04:07 PM 7/19/99 -0700, spober at manhattan.edu wrote:

>Since the main campus domain name server was up and has the correct DNS
>information, and since this situation corrected itself (apparently
>spontaneously) sometime late Sunday night or early Sunday morning, it's
>making me very curious about domain name propagation on the Internet.

Network Solutions distributes updated routing tables around midnight EST, 
thus it sounds as if there may have been an error propagated for a couple 
of days.  Figuring out what happened might be difficult -- they don't keep 
archives of the tables, unfortunately.  If you have a friend who operates a 
DNS server, who has a backup of the state of the server at that time, whose 
server tried to resolve the name, you might figure it out.  More likely, 
though, you'll just find that there was no entry at all.

DNS is becoming incredibly complex, especially as the new registrants have 
begun testing their systems.  One of my best friends is the CTO at Network 
Solutions -- his job has been described as running the most complex 
computing system ever created.

If I were you, I'd monitor more closely for a while and dig into the 
problem if it happens again.  But this sort of thing is going on relatively 
often on the net these days.  Even if the root servers are correct at Net. 
Solns., someone down the chain can mess things up temporarily.  It's a bit 
worrisome that your problem outlasted at least one update, however.

Nick