some users can't ping longtime established site

We have a site that has been up and running for over a year, and we have thousands of users, and at least hundreds of unique users every day. In the last week we have received a few calls from disparate customers around the country who have been using us regularly and all of a sudden have been unable to reach our site. They can neither ping our name nor our IP address (host not found).

I am well aware of the trouble that people can have when you move a site to a different IP address--how it takes a while to propagate across the net. The last time we did that was several months ago, and these users have been using us since then. Furthermore, those problems involved not being able to hit the name, not the address of the site.

Our servers are at a top-notch managed facility with a backbone connection to the net, and they're supposedly investigating our issue. We also contacted our ISP and they know of no reason any of our users would be having a problem. If it was just one or two, I'd say their own internal DNS servers or their ISP's were messed up, but we're getting more and more calls.

If anybody has the foggiest idea what could be causing this, that would be most helpful! Thanks...

Well me for one am not a magician so you would need to tell me your DNS name and IP address to give you a better answer if this is a public site thats not a big deal. what you might want to do is get on different looking glass servers as there may be a routing issue with some ISPs or one isp in the middle. the best thing you can do is request a trace route from the people that cant get to your site and then trace back to them. this would give you where the problem is.
To be more specific this comes from an ISP being multi-homed, it accepts routes via one peer and does not filter them when advertising its routes to the other peer via BGP causing an outage for anyone that traverses the peer that accepted the bad routes. The internet is asyncronous and you would need to trace both directions to find where the problem is. Usualy good ISP will put a 10% limit over the amount of agreed upon routes it will accept before BGP peering goes down on its own but smaller ISPs sometimes dont do this and leak unknowingly. unfortunately this will need to be resolved at an ISP level. I have also seen issues with MPLS tunnels on CEF enabled routers (namely the 7000 series) that will black hole traffic traversing the tunnel only for a certain IP address (at the ingress router) this is a pain to trouble shoot as the circuit and tunnel all look good and passes traffic except for a few destinned IP addresses. The only way to fix this is to reboot the router or shutting down the power to the VIP or line card (there are hidden commands in cisco IOS to do this). (This bug may have been fixed though) but again nothing you can do about it cause its at an ISP level other than tell them what router is dropping it from the trace. Man im getting way to talkative hehe!

Your posts just keep on getting longer and longer