Virtual Machine looses connectivity with no clear explanation

3 Jul 2017

      This is a multi-part message in MIME format.
--------------CA6FA101A23C7121C82BF9C5
Content-Type: text/plain; charset=utf-8; format=flowed
Content-Transfer-Encoding: 7bit

I have a rather strange issue which is affecting one of my last deployed 
Hypervisors. It is a CentOS 7 (not a oVirt Node) which runs only 3 
Virtual Machines.

One of these VMs have a reasonable output traffic at peaks (500 - 
700Mbps) and the hypervisor underneath is connected to the switch via a 
bonding (mode=2) which in turn creates bond0.XX interfaces which are 
connected to different bridges for each network. The VM in question is 
connected to bridge "ovirtmgmt".

When the problem happens the VM stops passing traffic and cannot reach 
even the router or other VMs in the same Layer 2. Seems the bridge stop 
passing traffic for that particular VM. Other VMs work fine since they 
were created. When this problem happens I just need to go to its Console 
and run a reboot (Ctrl-Alt-Del), don't even need to Power Off and Power 
On again using oVirt Engine.
I have even re-installed this VMs operating system from scratch but the 
problem persists. Have also changed the vNic mac address in the case 
(already check) of conflicted mac addresses somewhere in that Layer 2.

Last, my hypervisor machine (due a mistake) has been running with 
SElinux disabled, not sure if it could have anything to do with this 
behavior.

Anyway, anyone has ever seen any behavior like that ?

Thanks
Fernando

--------------CA6FA101A23C7121C82BF9C5
Content-Type: text/html; charset=utf-8
Content-Transfer-Encoding: 7bit

<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=utf-8">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <font face="arial, helvetica, sans-serif">I have a rather strange
      issue which is affecting one of my last deployed Hypervisors. It
      is a CentOS 7 (not a oVirt Node) which runs only 3 Virtual
      Machines.<br>
      <br>
      One of these VMs have a reasonable output traffic at peaks (500 -
      700Mbps) and the hypervisor underneath is connected to the switch
      via a bonding (mode=2) which in turn creates bond0.XX interfaces
      which are connected to different bridges for each network. The VM
      in question is connected to bridge "ovirtmgmt".<br>
      <br>
      When the problem happens the VM stops passing traffic and cannot
      reach even the router or other VMs in the same Layer 2. Seems the
      bridge stop passing traffic for that particular VM. Other VMs work
      fine since they were created. When this problem happens I just
      need to go to its Console and run a reboot (Ctrl-Alt-Del), don't
      even need to Power Off and Power On again using oVirt Engine.<br>
      I have even re-installed this VMs operating system from scratch
      but the problem persists. Have also changed the vNic mac address
      in the case (already check) of conflicted mac addresses somewhere
      in that Layer 2.<br>
      <br>
      Last, my hypervisor machine (due a mistake) has been running with
      SElinux disabled, not sure if it could have anything to do with
      this behavior.<br>
      <br>
      Anyway, anyone has ever seen any behavior like that ?<br>
      <br>
      Thanks<br>
      Fernando<br>
    </font>
  </body>
</html>

--------------CA6FA101A23C7121C82BF9C5--

FERNANDO FREDIANI

tags

participants (1)