I have two identical Dell PowerEdge T620 servers both server are running Esxi 5.5. The servers have dual Intel I350 network cards. I currently have each host only using one NIC interface, the other interface is not plugged in or configured. The server are setup in a test/lab environment connected to a unmanaged DLink DGS-1024D switch. When I cold reboot the host the NIC will start but then get disabled. Let me explain, from the host screen if I reboot the server hardware, the NIC interface connection light will come on. As the host comes up the NIC is then disable (no link light). I can not ping from the host or to the host's IP. In the same server box the is a second network card (Broadcom BCM5720) that is used for my virtual machines. This NIC does work even when the hosts NIC is not work (I can get to the virtual machines but not the host). I have adjusted the power settings of the host from "Balanced" to "Performance" but the problem still persists. Does anyone have any recommendations for this problem?
Hi
Welcome to the communities.
There might be many thins , please go through below
change LAN cable,re install driver , check HCL details
Jayden56 thanks for the help. I tried the obvious things like network cable and network switch. The NIC is listed as Intel Corporation I350 Gigabit Network Connection and the driver is: igb.
The Device State is: 8086:1521 1028:04cf vmkernel vmnic0 . Interesting when I look up those numbers on the compatibility guide they are not listed. The server was purchased directly from Dell with VSphere Essientials Kit installed. I would have thought they would have used a compatible NIC. I am going to try change over to the Broadcom secondary NIC to test. I will post my results.
Are you using the Dell VMware ESXi 5.5 image? This includes some additional drivers that might get you going with this NIC.
Thanks, yes I am using the Dell VMware ESXi 5.5 image. I only see one driver listed which is NET_IGB.V00. According to the VMware compatibility guide I am using the latest driver for this device - version: 4.2.16.8.
Can you do an esxcli network nic list via SSH?
Name PCI Device Driver Link Speed Duplex MAC Address MTU Description
------ ------------- ------ ---- ----- ------ ----------------- ---- -------------------------------------------------------
vmnic0 0000:006:00.0 igb Up 1000 Full f0:1f:af:e1:b3:81 1500 Intel Corporation I350 Gigabit Network Connection
vmnic1 0000:006:00.1 igb Down 0 Half f0:1f:af:e1:b3:82 1500 Intel Corporation I350 Gigabit Network Connection
vmnic2 0000:001:00.0 tg3 Up 1000 Full 00:0a:f7:3e:5f:28 1500 Broadcom Corporation NetXtreme BCM5720 Gigabit Ethernet
vmnic3 0000:001:00.1 tg3 Up 1000 Full 00:0a:f7:3e:5f:29 1500 Broadcom Corporation NetXtreme BCM5720 Gigabit Ethernet
vmnic4 0000:003:00.0 tg3 Up 1000 Full 00:0a:f7:3e:5e:9c 1500 Broadcom Corporation NetXtreme BCM5720 Gigabit Ethernet
vmnic5 0000:003:00.1 tg3 Up 1000 Full 00:0a:f7:3e:5e:9d 1500 Broadcom Corporation NetXtreme BCM5720 Gigabit Ethernet
So the pNIC is identified as vmnic0 and is successfully loaded. Are you sure that your switch doesn't have anything to do with this weird behaviour? Perhaps some kind of STP (Spanning-Tree) related issue? The pNIC seems online and connected at 1 gb/s.
Yes it is vmnic0 that I am dealing with. I have tried two different switches a DLink DGS-1024G and a Cisco SG100D-08 both give me the same result. Right now the NIC is UP and Online but by tomorrow morning it will be off again. When it is off-line if I unplug the cable from one switch and plug it into a different switch the link light on vmnic0 will not come on. If this was a spanning-tree issue I would expect the link light to at least come when plugged into a different switch. If I reboot the host, vmnic0 still does not work, but if I power the host completely off and then on again the vmnic0 will work for a while usually.
If you could, configure a second VMkernel enabled for management on your other links. When you see this issue happening with vmnic0, could you post the vmkernel.log (located in /var/log/vmkernel.log) here? That might help.
Will do. Thanks again for all your help
There is a more recent igb driver available from VMware, namely version 5.0.5.1-1290233:
https://my.vmware.com/group/vmware/details?downloadGroup=DT-ESXI55-INTEL-IGB-5051&productId=353
Update the driver with this offline bundle and see if your issue persists. Also make sure the firmware of the NIC is up to date.
If the it still occurs check the logfile /var/log/vmkernel.log.
OK, I updated the driver software for the NIC and the problem was still there. I did some poking around and found the on-board NIC set to "Auto Negotiate". I changed the setting to 1000 Mb, Full Duplex. Since making this change I have had no problem with the NIC turning off.
Hopefully this helps anyone else who is experiencing the same problem.