locked
Dropping packets when pinging VM's on specific hyper-v host!? RRS feed

  • Question

  • I have 3 hyper-v server machines.  On one of them I have 3 VM's and on the other I have 9 VM's.  One of my clients, all of the sudden started dropping packets randomly when continuously pinging any of the guests but not the host itself on the host with 3 VM's on it. ???

    Nobody else seems to be having these problems.  I tried updating the Windows 7 x64 client's NIC drivers, removing them, doing the netsh reset stuff, rebooting, updating windows, restarting the VM's, etc and to no avail!?  My personal computer (Vista x64) and other computers on the network have no trouble, just this single computer.

    My question is:  Why would this single client drop packets to only the guests on this specific hyper-v host when nothing has changed on either side all of the sudden?  The client machine can ping anything else, including other vm's on other hyper-v servers, no problem.

    Any ideas out there?

    Thursday, May 17, 2012 10:29 PM

Answers

  • I checked those things and could find nothing wrong.  However, others (after much prodding) indicated they were having random issues with that same VM but in different ways, though they could ping it just fine.

    However, just for giggles, I decided to reboot the host and all of the VM's on that box.  Funny thing is, all the problems went away!?  The client could then ping the VM's just fine and the other odd issues went away as well.  This definitely was a Hyper-V issue and a reboot resolved it.  I still don't know what it was but it's resolved.

    Thanks for your replies!  If the situation changes or the problem comes back I'll be right back on here and we can dig into it some more.

    • Marked as answer by SCADirector Sunday, May 20, 2012 6:52 PM
    Sunday, May 20, 2012 6:51 PM

All replies

  • Hi,

    It seems that this is a network issue instead of a Hyper-V issue, you need to use network monitor to capture the network communication. It is recommended that you perform the further research in networking corresponding community so that you can get the most qualified pool of response. Thanks for your understanding.

    Sunday, May 20, 2012 1:55 PM
  • Let me start by recap your question: you have 1 hyper-v host that hosting 3 vm. There is one specific PC in the network fail to ping one vm.

    Investigation: The hyper-v network hyper-v host is to assign one NIC to be virtual switch. All VM connect to external network switch for communication outside the host. By default, Hyper-V host has no capability to act as firewall between VM and external network. You should consider from VM to PC and something in between e.g. router, ACL on swtich.  It is worst to perform basic investigation in below.

    - At client PC, transfer IP address to other PC and reproduce the problem. If, it is working, then, the problem is on PC site. You need to check Windows Firewall, Anti-virus, eventlog.  


    ___________________________________________________ Naruphon blog: http://www.vm360degree.com

    Sunday, May 20, 2012 3:12 PM
  • I checked those things and could find nothing wrong.  However, others (after much prodding) indicated they were having random issues with that same VM but in different ways, though they could ping it just fine.

    However, just for giggles, I decided to reboot the host and all of the VM's on that box.  Funny thing is, all the problems went away!?  The client could then ping the VM's just fine and the other odd issues went away as well.  This definitely was a Hyper-V issue and a reboot resolved it.  I still don't know what it was but it's resolved.

    Thanks for your replies!  If the situation changes or the problem comes back I'll be right back on here and we can dig into it some more.

    • Marked as answer by SCADirector Sunday, May 20, 2012 6:52 PM
    Sunday, May 20, 2012 6:51 PM
  • I came across this post after having the same lost packets from ping responses on 2 VMs that are on a 5-node, 5-VM Hyper-V Cluster.  I found that 2 VMS were constantly dropping packets...down to a 40-50% reponse rate.  I found that 1 VM was causing the issue for both VMs.  Since the offending VM was a test VM I deleted it, rebuilt the VM from the same template and brought it back into service.  All is good now on both VMs... no dropped packets on either VM.  What I didn't do was to move the VM from the cluster over to a Standalone host to see if I had the same issues and if the packet dropped would stop on the second VM.

    Tuesday, June 19, 2012 11:08 PM