none
destination host unreachable

    Question

  • I have a weird issue....I think.

    I just noticed that on my servers (all HP and setup by someone else), if I would ping an IP address of a downed machine or an IP not in use all in the same subnet, I get a message of "Reply from 10.10.10.20: Destination host unreachable and the server IP that I am pinging from is the 10.10.10.20.

    From my experience, I should be getting a "request timed out message" only if it's on the same subnet.  Anyone know why I'm getting the Destination host unreachable message when I'm not trying to reach a different subnet and why is it coming from my machine?

    Friday, January 21, 2011 4:24 PM

Answers

  • The response "Reply from 10.10.10.20: Destination host unreachable." actually is the correct one.  When pinging a host on your own subnet the sender will first send an ARP request for the MAC address of the associated IP address.  Since the host does not exist, the sending station is reporting to you that the destination host is unreachable because the ARP request was not completed.  The actual ICMP echo was never sent which is why you aren't getting a "request timed out" message.  You can't send a layer 3 packet if you can't resolve the layer 2 address of the host (or the gateway if the destination is on another network). 

    The reason why you would not get this same message if you were pinging hosts on a different subnet is that the sender would ARP for the gateway address since the target is on a different subnet.  The ARP completes properly and the ICMP echo is sent but the timeout value is reached since the destination host never responded with an echo-reply.

    It seems that Win7/2008 just gives more specific failure reasons now than Windows XP does.  This is a good thing in my estimation.


    Matt W. CCNP, CCDA, CCNA-S, RHCT, MCSE, MCSA, MCP+I, A+
    • Marked as answer by Tiger Li Wednesday, January 26, 2011 9:25 AM
    Tuesday, January 25, 2011 3:03 PM

All replies

  • Probably because the ARP request was not completed successfully so the ICMP echo was never actually sent.
    Matt W. CCNP, CCDA, CCNA-S, RHCT, MCSE, MCSA, MCP+I, A+
    Sunday, January 23, 2011 10:21 PM
  • You are correct...the expected response should be "Request timed out".  I have seen this behavior in Win 7...don't recall if this is also the case in 2008 as well.

    If you want to determine if the packets actually left the comptuer, run a packet capture to take a look.  Feel free to let us know the results.


    Visit: anITKB.com, an IT Knowledge Base.
    Monday, January 24, 2011 1:47 AM
  • Hi,

     

    Thanks for posing here.

     

    In addition ,this seem a most possible incorrect route setting issue on host , like default gateway setting.

    Could you verify the route table on this server and post here for further investigation ?

     

    Thanks.

     

    Tiger Li


    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.
    Monday, January 24, 2011 8:02 AM
  • Hi,

     

    Thanks for posing here.

     

    In addition ,this seem a most possible incorrect route setting issue on host , like default gateway setting.

    Could you verify the route table on this server and post here for further investigation ?

     

    Thanks.

     

    Tiger Li


    Please remember to click “Mark as Answer” on the post that helps you, and to click “Unmark as Answer” if a marked post does not actually answer your question. This can be beneficial to other community members reading the thread.
    Monday, January 24, 2011 8:02 AM
  • I verified that the default gatway is correct but it still shouldn't matter because the traffic is local to the subnet.

    Here's the output for the route print command.  Disregard the IP that was in the first post. 

    IPv4 Route Table
    ===========================================================================
    Active Routes:
    Network Destination        Netmask          Gateway       Interface  Metric
              0.0.0.0          0.0.0.0      10.65.235.1     10.65.235.20    266
          10.65.235.0    255.255.255.0         On-link      10.65.235.20    266
         10.65.235.20  255.255.255.255         On-link      10.65.235.20    266
        10.65.235.255  255.255.255.255         On-link      10.65.235.20    266
            127.0.0.0        255.0.0.0         On-link         127.0.0.1    306
            127.0.0.1  255.255.255.255         On-link         127.0.0.1    306
      127.255.255.255  255.255.255.255         On-link         127.0.0.1    306
            224.0.0.0        240.0.0.0         On-link         127.0.0.1    306
            224.0.0.0        240.0.0.0         On-link      10.65.235.20    266
      255.255.255.255  255.255.255.255         On-link         127.0.0.1    306
      255.255.255.255  255.255.255.255         On-link      10.65.235.20    266
    ===========================================================================
    Persistent Routes:
      Network Address          Netmask  Gateway Address  Metric
              0.0.0.0          0.0.0.0      10.65.235.1  Default
    ===========================================================================

    IPv6 Route Table
    ===========================================================================
    Active Routes:
     If Metric Network Destination      Gateway
      1    306 ::1/128                  On-link
     10    266 fe80::/64                On-link
     10    266 fe80::ed93:d3da:c363:6700/128
                                        On-link
      1    306 ff00::/8                 On-link
     10    266 ff00::/8                 On-link

    Monday, January 24, 2011 8:24 PM
  • I checked on 2008 and the behavior is the same as Windows 7.  I know that there were modifications to the TCP/IP stack after Win XP, but I have not had a chance to research this in detail or perform a PCAP to verify the contents of the ICMP packets on the network.

    No need to look at the routing table for layer 2 packet delivery.


    Visit: anITKB.com, an IT Knowledge Base.
    Tuesday, January 25, 2011 12:59 AM
  • The response "Reply from 10.10.10.20: Destination host unreachable." actually is the correct one.  When pinging a host on your own subnet the sender will first send an ARP request for the MAC address of the associated IP address.  Since the host does not exist, the sending station is reporting to you that the destination host is unreachable because the ARP request was not completed.  The actual ICMP echo was never sent which is why you aren't getting a "request timed out" message.  You can't send a layer 3 packet if you can't resolve the layer 2 address of the host (or the gateway if the destination is on another network). 

    The reason why you would not get this same message if you were pinging hosts on a different subnet is that the sender would ARP for the gateway address since the target is on a different subnet.  The ARP completes properly and the ICMP echo is sent but the timeout value is reached since the destination host never responded with an echo-reply.

    It seems that Win7/2008 just gives more specific failure reasons now than Windows XP does.  This is a good thing in my estimation.


    Matt W. CCNP, CCDA, CCNA-S, RHCT, MCSE, MCSA, MCP+I, A+
    • Marked as answer by Tiger Li Wednesday, January 26, 2011 9:25 AM
    Tuesday, January 25, 2011 3:03 PM
  • I have seen others post this similar question, and I beleive that he confusion is because prior to Win7/2008, the response that a user would see when trying to ping a host on the same segment that does not respond would be "Request Timed Out", not Destination Host Unreachable.  It does not appear that rikity is confused about layer 2/layer 3.  His orginal post clearly identifies that the target systems are on the same segment.


    Visit: anITKB.com, an IT Knowledge Base.
    Tuesday, January 25, 2011 3:32 PM
  • So are you saying that this is totally normal for Windows 2008 to give a Destination Host Unreachable message when pinging and IP in the same subnet?
    Tuesday, January 25, 2011 8:04 PM
  • yes it would appear so
    Matt W. CCNP, CCDA, CCNA-S, RHCT, MCSE, MCSA, MCP+I, A+
    Tuesday, January 25, 2011 8:22 PM
  • I had a different situation that I "THOUGHT" was windows 2008 related.  Turns out that the virtual machines I was using were not being given UNIQUE MAC addresses.  It was the comments above about ARP that made me realize what was happening.  I made manual settings and then PING started working because ARP could do its job.

    Oddly enough... RDP worked to each of these virtual machines but that traffice went through a gateway and thus the local subnet didn't seem to be an issue.

    Just thought I would share.


    Monitoring is the art of looking for something you don't want to find.
    Wednesday, March 02, 2011 8:21 PM
  • Since you bring up the point about unique MAC addresses in the virtual environment, if you are using Virtual PC, here is an overview on how to modify the settings.  This usually happens if you simply copy the configuration file from an existing set.


    How to Edit the MAC Address for Microsoft Virtual PC Guests
    http://www.anitkb.com/2010/05/how-to-edit-mac-address-for-microsoft.html

     


    Visit: anITKB.com, an IT Knowledge Base.
    Wednesday, March 02, 2011 9:53 PM
  • Hi,

     

    I'm having the same issue (Destination host unreachable) when trying to ping an internal platform I need to access.  Thing is, I'm using Windows 7's, Windows XPMode (ie. virtual machine).  It's strange because I CAN ping the SAME IP address from my host machine (ie. Windows 7 OS) but get this issue when trying to ping from the XP mode (ie . virtual machine) only.

     

    Any ideas guys?

     

    PS - I NEED to be able to contact the platform from XP Mode because the test case I am trying to run does NOT operate on Win 7.

    Friday, August 26, 2011 9:05 AM
  • If 'destination host unreachable' is "correct" how do you explain this from running a ping to a machine being rebooted on the same segment from a Windows 7 machine?

    Pinging xxx.yyy.zzz.com [nnn.mmm.75.50] with 32 bytes of data:
    Request timed out.
    Request timed out.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Request timed out.
    Request timed out.
    Request timed out.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    .
    .
    .
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Request timed out.
    Reply from nnn.mmm.75.131: Destination host unreachable.
    .
    .
    .
    Reply from nnn.mmm.75.131: Destination host unreachable.
    Reply from nnn.mmm.75.50: bytes=32 time=999ms TTL=128
    Reply from nnn.mmm.75.50: bytes=32 time<1ms TTL=128
    Reply from nnn.mmm.75.50: bytes=32 time<1ms TTL=128
    Reply from nnn.mmm.75.50: bytes=32 time<1ms TTL=128

    .75.131  is the machine running the ping,  .75.50  is the target. Windows 7 seems rather conflicted to me about the situation  :) .

    Friday, June 22, 2012 1:42 AM
  • @GRL570810: it's simply explained - when you were getting "Request timed out" the destination was responding to ARP requests but not to ICMP requests, when it was respondiing to neither, you got "Destination host unreachable". During boot up various processes may come and go as the BIOS and OS sort things out, once the boot up was complete both protocols were responding and the ping completed.
    • Edited by ew1001 Thursday, June 28, 2012 10:35 AM
    Thursday, June 28, 2012 8:34 AM
  • Just adding what solved my case...

    I have DSL and a separate router setup as a separate network, e.g. DSL router is 192.168.2 and the other router 192.168.0. One day I came home to all the machines powered off, except for the routers and other things; a lengthy power failure caused the UPS to run out of juice... after restarting the machines (linux and windows/7) on the 192.168.0 network I got the "Destination Host Unreachable" response from one box to the other. Rebooting the router solved it.

    Thursday, August 23, 2012 3:58 PM