Just want to follow up on my first question. It turned out to be a misbehaving interface, i.e. Layer 1 issue. You do get a reply with a broadcast Ping from each host in the subnet. However, in spite of the one router interface giving me issues, I'm Still trying to TRACE the logic as to how a second router can elicit a PING reply from the first router when Router1 cannot initiate the PING, in spite of the fact that both have each other's ARP table Aware of each other. If only the "debug ip icmp" would give you an Error output when the PINGs don't succeed. Thanks. Elmer P.S. Kevin, I wish I had vocabulary like yours :->
-----Original Message----- From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]] On Behalf Of Kevin Cullimore Sent: Sunday, August 18, 2002 3:08 PM To: [EMAIL PROTECTED] Subject: Re: Simple (silly) question on PING [7:51580] ----- Original Message ----- From: "cebuano" To: Sent: 17 August 2002 11:35 pm Subject: Simple (silly) question on PING [7:51580] > Hi all, > Just two simple but annoying PING questions. > 1. Why is it that on a broadcast medium, you can issue ping > 224.0.0.5 and have all OSPF routers respond (according to CCO), but when > I do a simple ping 192.168.10.255 on the subnet, no replies are seen > from all the interfaces on this subnet? I know you'll say my brain is > getting fried from too much "rack exposure". This is one of those cases that underscore the extent to which communication protocol specifications sometimes define a range of acceptable behavior in the face of a given set of conditions rather than a single acceptable option. In some cases, the RFCs/Standards don't provide a reccomendation for how a given implementation should behave, leading to real-world interoperability issues. A more obvious case where these considerations matter involve the inexplicably persistent notion that distinct implementations of a given standard should behave identically given identical circumstances. In this case, whether or not a given icmp/ip implementation responds to echo requests addressed to a layer 3 broadcast address is left up to the vendor (which used to provide a quick-n-dirty way of performing simplistic os fingerprinting within a given broadcast domain), based on the use of the may keyword when describing within RFC 1122 when describing the receiving host's behavior during that situation. The key here is that the guidelines covering behavior in response to received multicast & broadcast traffic are separate, allowing for distinct behavior, which may, in turn, reflect different needs/goals to be addressed when dealing with the two different (though conceptually related) types of traffic. > 2. Using a crossover to connect two Ethernet interfaces, I can ping > say RtrA's e0 from RtrB, but can't ping RtrB's e0 from RtrA. I know some > of you on the list have seen this before and have had a really > crystal-clear explanation for this. > > TIA, > Elmer Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=51612&t=51580 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
