3

I recently did an exam where one of the questions stated that you can run ping 127.0.0.1 on a computer to verify if its network card is working.

I have already read on multiple sources that this actually only guarantees that the networking software is installed properly. For example, the book "Network Fundamentals, CCNA Exploration Companion Guide" states that:

"A response from 127.0.0.1 indicates that IP os properly installed on the host. (..) This response is not, however, an indication that the addresses, masks or gateways are properly configured."

Wikipedia also says that "Using the loopback interface bypasses local network interface hardware.", but provides no sources.

The problem is, I need an official source saying that pinging 127.0.0.1 does not hit the network card.

RFC 5735 says this:

127.0.0.0/8 - This block is assigned for use as the Internet host loopback address. A datagram sent by a higher-level protocol to an address anywhere within this block loops back inside the host. This is ordinarily implemented using only 127.0.0.1/32 for loopback. As described in [RFC1122], Section 3.2.1.3, addresses within the entire 127.0.0.0/8 block do not legitimately appear on any network anywhere.

It's not clear when the "loop back" occurs.

Any help? Thanks.

Community
  • 1
Schrute
  • 181
  • 1
  • 3
  • 1
    This is so blatantly obvious that I would be rather shocked that someone would get it wrong. Consider a computer with _no_ network cards. 127.0.0.1 still works. – Michael Hampton Apr 26 '15 at 16:54
  • @MichaelHampton why is it obvious? I don't see any document stating that it should work (or shouldn't). – Schrute Apr 26 '15 at 17:16
  • @MichaelHampton (I'm just trying to understand it so I can explain it to the exam board) – Schrute Apr 26 '15 at 17:18
  • 2
    You could look at [the source code](https://github.com/torvalds/linux/blob/master/drivers/net/loopback.c), which makes it very obvious that loopback is a virtual interface not connected to any hardware. – Michael Hampton Apr 26 '15 at 17:21
  • The overall approach is described [for IPv6](https://tools.ietf.org/html/rfc2373#section-2.5.3) but I'm not positive if there's some rfc/std where it's spelled out that way for IPv4. – Håkan Lindqvist Apr 26 '15 at 17:25
  • 2
    This is one of those type of exam questions that is too easy to "overthink". For example, two virtual machines on the same host could in theory do "networking" without any network card; so, in short, you won't find this: "I need an official source saying that pinging 127.0.0.1 does not hit the network card." The answer to the question is simply (probably) "yes", as it is a useful first diagnostic step that is "necessary but not sufficient" to determine if networking is configured. – michael Apr 26 '15 at 18:08
  • @MichaelHampton There's no need to be unprofessional. Its a legit question that does not have an 'official' documented answer that I can find either. – Nicholas Terry Jan 03 '18 at 17:50
  • @nterry What are you talking about? – Michael Hampton Jan 03 '18 at 17:55
  • 1
    @MichaelHampton Your first comment. Although this is an older question, I was simply expressing my surprise at how abrasive you were. And i quote: 'This is so blatantly obvious that I would be rather shocked that someone would get it wrong. Consider a computer with no network cards. 127.0.0.1 still works' – Nicholas Terry Jan 09 '18 at 19:10

0 Answers0