PowerEdge R720 VM's cannot ping VM host

Hi,

The 4 ports that go to the lan, 2 to each switch, are the switches on the same subnet? Do they connect to each other? Did you replace the cables when you went from Cisco to Dell? Was this all working when you used Citrix?

This server has always given us an issue.  The old Citrix is on an HP Blade server that we are getting rid of.  The new Citrix servers were created in VMWare and are currently on the other 2 hosts.  I only have one VM on vm8 for testing purposes.  Everything is on the same subnet and we have the 4 Dell switches stacked.  I have not replace the cables.

It could be a cable issue, or a NIC issue, with a lot of the other hardware and software being changed it probably is not one of those devices.

One thing you might want to try is take the network redundancy out of the equation. Try it with everything only connected to switchA, then only connected to switchB. If it works on one side but not the other, you can narrow your focus to things like cables or switch ports in that stack. If both singles work, but not when redundant, it's more likely a switching/routing issue where not all paths between sets of adapters are valid.

Since there are 4 vnics on this server and we have 3 switches, I started moving the connections around.  We have 2 vnics on switchC and 2 on switchB.  I moved the 2 from switchC to switchB (all 4 on switchB).  Then I took the original 2 on switchB and moved them to switchA.  Then I took the 2 on switchB and moved them to switchC.  Then the 2 on switchA to switchC.  This was just about every possible combination.  I still had issues.  I was even checking the system when only 2 vnics were online and still had issues.  By testing when only 2 vnics were tested that ruled out a cable issue unless I have 2 or more bad cables.  I also tried different ports on the switches.   It doesn't appear to be switch or cable problem.  When I checked the settings in vSphere on a 3 hosts, every settings is exactly the same.  I used the iDRAC and logged into each server and every setting is the same (BIOS, firmware, drivers, etc.)

Now as I was typing this, I tried to log into iDRAC on the server in question.  I entered my username/password and it spun on the verifying credentials screen for several minutes.  I used iDRAC to check a different server and got right in.  I retried the first server again and this time, now I get a "is not available".  And in my vSphere, that host has an isolation error reaching the server.  

Since I cannot get into iDRAC and having issues even getting into the server, could it be a hardware issue with the server????  And how do I even test the server????

This is happening across multiple servers, right? It is unlikely that all of them have the same NIC issue. You could try swapping to a different NIC and see if that works better.

No, this is isolated to one R720 server.  I removed the Broadcom 5719 nic adapter and only used the 5720 onboard, with the same results.  Put the 5720 back it and only used that nic (nothing in the 5719 nic ports) with the same results.  Yesterday, I tried to using iDRAC to get into the server and it hung on the Verifying Credentials screen after entering the username/password.  I closed the window and tried to go back in and I got a browser message stating that the URL was unavailable.  I tried iDRAC on my other 2 R720's and got right in.  About 10 minutes later, I tried to go back in with iDRAC and got in.  I am thinking that there is a hardware issue, but I am having a hard time diagnosing it.  I verified every setting between my 3 R720's and everything is the same. (Same BIOS, firmware, and drivers).  Even vmWare support has been checking my servers and they are totally puzzled.  They determined that it cannot be a cable or physical switch issue, since a virtual machine is unable to ping the host that is residing on.  When a VM pings it host, it only uses its vSwitch.  It doesn't go out to the physical switch.  There is an issue with the tg3 driver (Dell and vmWare recognize this).  Back in October, I upgraded to the tg3 driver that does not have the packet loss issue.  Also, the work around is to disable the NetQueue, which I did last night.  This also did not help.  The server is under warranty until July 2017.

Can you boot to our Live image and see if it has issues pinging in a different OS? http://www.dell.com/support/contents/us/en/19/article/Product-Support/Self-support-Knowledgebase/enterprise-resource-center/Enterprise-Tools/support-live-image

Can you get a DSET report?