>>>>> "Rich" == Rich Shepard <[email protected]> writes:
Rich> While out of the state last week I was not able to use ssh to Rich> access my office server. The hotel replaced their wifi with Rich> ethernet (yay, team!) so there was no problem obtaining an IP Rich> address from their provider's server. However, packets would go Rich> no further than a Frontier Communications server in Beaverton. Rich> I would like to understand why I could ping the server (mail Rich> dot appl-ecosys dot com) by name or IP address (which I obtained Rich> using 'host' since Frontier changes it several times each evening Rich> and night), but traceroute and ssh failed to reach the same Rich> destination. [...] Rich> It is curious that traceroute fails at hop 8 for the name but Rich> at hop 5 for the IP address. N.B. The two IP addresses are the Rich> result of my saving the output on two different days and Frontier Rich> had changed it between attempts. Rich> My interpretation is that this is a telco issue. Is that Rich> correct? If so, is there anything I can do should this problem Rich> re-appear the next time I'm away from the office? Ping and ordinary traceroute is an inadequate test. ICMP ping requests are sometimes dropped. You should also test with tcptraceroute, which disguises itself as a regular TCP connection and is much less likely to be dropped. You can try port 80 or port 22, and that should tell you where the packets get lost, e.g.: $ tcptraceroute <ipaddr> <port> One possibility is that your IP address was stale and pointing at someone different than your server. There are many other possiblities. -- Russell Senior, President [email protected] _______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
