On my Win 7 Pro 64 bit, I use VPN to connect to a couple of different clients. I have now experienced this issue on two clients VPN's.
There is a newsgroup server nntp://forums.novell.com that I want to check messages on. While I am waiting, I swap over in Thunderbird to go check.
But using the AnyConnect client at one server, sometimes (so I assume one node of VPN cluster servers) I get an error connecting. tracert forums.novell.com returns a General Failure before looking at the first hop.
C:\Users\geoffc>tracert forums.novell.com
Tracing route to forums.novell.com [130.57.5.50]
over a maximum of 30 hops:
1 General failure.
Trace complete.
DNS is not the issue as it shows the correct IP address.
Then on a second client, I switched from using Anyconnect (and NNTP there worked) to using the thick Cisco VPN client (64 bit, 5.0.07.0290) and suddenly it stopped working again.
But here is where it gets funny. I was showing this one of our network guys, and I tracert'ed the IP and it failed. Then for fun, I tried IP + 1. (it ends in .50 so I tried .49 and .51) and then suddenly it started working.
I got disconnected by a network burp. When I reconnected, this repeated itself.
And when I tried again, same error, but this time I just did a tracert to the IP (skipping DNS resolution) and it started to work:
C:\Users\geoffc>tracert 130.57.5.50
Tracing route to 130.57.5.50 over a maximum of 30 hops
1 42 ms 42 ms 42 ms l3-umvpn-arbl.r-arbl.acme.com
2 44 ms 42 ms 41 ms l3-barb-rarb.r-bin.acme.com
3 42 ms 42 ms 42 ms l3-barb-bseb-2.r-bin-s.com
4 48 ms 41 ms 49 ms v-bin-seb-inet-aa2.acme.com
5 ^C
And then it suddenly is working.
We are thinking of possible scenarios...
A router with a bad ARP cache? But it dies before even making the first hop in tracert.
But more than anything else, what possible mechanism would trying to connect to another IP nearby suddenly make it work?
