VCP 5 – Objective 6.2 – Perform Basic vSphere Network Troubleshooting

Verify network configuration

There are a couple of ways you can verify the network configuration from within the vSphere client.  One of my favorite ways is through the Maps tab.  Here you can specify to view Host to Network as well as VM to Network and get a visual view of which hosts and which VMs are attached to the networks.  As well in the Networking inventory view you can get some very nice lists showing the VMs and Hosts that are connected to different networks and port groups.
 
Verify a given virtual machine is configured with the correct network resources
 
There are a number of troubleshooting steps that you can take to verify VM network connectivity.
  • Ensure that the network associated with your VMs NICs actually exists and is spelled correctly within your virtual switches and that the connected box is checked.
  • Ensure that the VM has no underlying storage problems or is not under contention.
  • Verify that networking from within your guest OS is configured properly.
  • If the VM was converted from a physical machine, verify that there are no hidden network adapters.
  • Verify that the vSwitch has enough ports to support the VM.
Troubleshoot virtual switch and port group configuration issues
 
Just as mentioned above be sure that the port groups exists across all hosts and that they are named exactly the same.  Also, know if you have ports available.  Although the total usable ports on a vDS or VSS is 4088 (even though it actually has 4096 – 8 are reserved) you can only have a maximum of 1016 active ports per switch.  If using a vDS be aware that you can also only have a maximum of 30000 ports per vCenter and further more have a look at your port binding options.   If you have selected 'Static Binding' then a port is assigned to a VM when it is connected to the port group, thus if you have no more ports available on your port group you cannot connect VMs to it.  If you have selected ephemeral ports are created and removed dynamically during power-on, power-off, connect and disconnect operations.  Dynamic portgroups are no longer available in vSphere 5.
 
Just as with any other network troubleshooting, if you are using VLAN's be sure that you have selected the correct port group and that it is tagging on the proper VLAN.
 
Troubleshoot physical network adapter configuration issues
 
In the physical nic you can look at speed and duplex mismatch settings.  Certainly drivers as well.  Depending on your teaming policy selection you may need to enable etherchannel or link aggregation on your switches.  You can also check VLAN information on the ports that the NICs are connected to.   I'm not sure what else that you could check at this point.  Normally there isn't a whole lot that can go wrong here.
 
Identify the root cause of a network issue based on troubleshooting information
 
I would normally start at finding the root cause of a network issue at the the VM.  If a VM has lost network connectivity, first thing I would check is to see if other VMs on the host or the host itself as lost network connectivity.  You can use the vmkping -D command to ping out through all of your vmkernel port groups in order to determine if traffic is getting out.  Simply combine all of the other points in order to determine where the network failure has occurred and go from there.  As well in ESXTOP have a look at DRPTX (transmit packets dropped) and DRPRX (Received packets dropped) to help troubleshoot as well.  ESXTOP can be very beneficial in troubleshooting as you will be able to see what VM is mapped to what uplink

Leave a Reply

Your email address will not be published. Required fields are marked *