ESXi 6.5 U1 - host and vms are not responding

[ivaylodyakov]

Im running ESXi 6.5 U1 on Fujitsu PRIMERGY rx2530 m4 with the latest Fujitsu custom image for PRIMERGY servers.  The problem is that, one or two times of day the vms and host are no responding and i need to restart the host locally when i connect to server with monitor and keyboard, because i not access of it. Is there someone with same problem or similar?

[jhague]

Are you able to ping the host? Have you tried looking at the vmkernel log while connected locally while host is not responding?

[ivaylodyakov]

While host is not responding, i dont have ping to host and vms. At same time there are access to other devices and the problem is not from network equipment.

Im not tried looking at the vmkernel log while host is not responding, because when server is down, it stop the work of 200 employees and im in a hurry to restart the host.

[hussainbte]

Can you share the Network configuration of the ESXi host. the vSwitch details and physical network configuration

[ivaylodyakov]

There are 6 physical NICs and 3 vswitches.

vswitch1 - management and first local network - vmnic0 - 172.10.30.0/24; vmk0

vswitch2 - second local network - vmnic1 192.168.10.0/24; vmk1

vswitch3 - nfs storage communication - vmnic2-vmnic5 - 10.10.10.0/24; vmk2

[jhague]

I think more info is required, that server is certified with 6.5 U1 so will have been through the testing process and everything should work. If you can't examine the vmkernel log at the time you can have a look at entries for the time the server failed once it is back up. If you are able to post them up here someone may be able to help further.

[hussainbte]

what is phyiscal switch configuration? are you using any teaming ?

when the host goes not responding are you loosing the ping to ESXI IP and VMs.?

share the nic details and driver and firmware used.

command

ethtool -i vmnic0

[ivaylodyakov]

This are screenshots from vmkernel log while the host is not responding.

https://drive.google.com/open?id=0B5UgzPsgN8nPbHlkak94VGd0dkk

https://drive.google.com/open?id=0B5UgzPsgN8nPT1I0LWJJTF9Hd1U

https://drive.google.com/open?id=0B5UgzPsgN8nPLTFUdmo5MVktdTg

https://drive.google.com/open?id=0B5UgzPsgN8nPZFBoSkFoSXNIV0k

[ivaylodyakov]

When the host is goes not responding i lose the ping to ESXI ip and vms.

There are 3 different hardware network devices for three networks without any vlans configuration.

Name    PCI          Driver      Link Speed      Duplex MAC Address       MTU                                         Description

vmnic0  0000:01:00.0 igbn        Up   1000Mbps   Full   90:1b:0e:d4:17:e4 1500                                        Intel Corporation I350 Gigabit Network Connection

vmnic1  0000:01:00.1 igbn        Up   1000Mbps   Full   90:1b:0e:d4:17:e5 1500                                        Intel Corporation I350 Gigabit Network Connection

vmnic10 0000:86:00.0 i40en       Down 0Mbps      Half   3c:fd:fe:a7:de:c8 1500                                        Intel(R) Ethernet Controller X710 for 10GbE SFP+

vmnic11 0000:86:00.1 i40en       Down 0Mbps      Half   3c:fd:fe:a7:de:c9 1500                                        Intel(R) Ethernet Controller X710 for 10GbE SFP+

vmnic12 0000:86:00.2 i40en       Up   10000Mbps  Full   3c:fd:fe:a7:de:ca 1500                                        Intel(R) Ethernet Controller X710 for 10GbE SFP+

vmnic13 0000:86:00.3 i40en       Up   10000Mbps  Full   3c:fd:fe:a7:de:cb 1500                                        Intel(R) Ethernet Controller X710 for 10GbE SFP+

vmnic2  0000:3d:00.0 i40en       Up   1000Mbps   Full   90:1b:0e:a9:b5:39 1500                                        Intel(R) Ethernet Connection X722 for 1GbE

vmnic3  0000:3d:00.1 i40en       Down 0Mbps      Half   90:1b:0e:a9:b5:3a 1500                                        Intel(R) Ethernet Connection X722 for 1GbE

vmnic4  0000:3d:00.2 i40en       Down 0Mbps      Half   90:1b:0e:a9:b5:3b 1500                                        Intel(R) Ethernet Connection X722 for 1GbE

vmnic5  0000:3d:00.3 i40en       Down 0Mbps      Half   90:1b:0e:a9:b5:3c 1500                                        Intel(R) Ethernet Connection X722 for 1GbE

vmnic6  0000:18:00.0 i40en       Down 0Mbps      Half   3c:fd:fe:a7:df:38 1500                                        Intel(R) Ethernet Controller X710 for 10GbE SFP+

vmnic7  0000:18:00.1 i40en       Down 0Mbps      Half   3c:fd:fe:a7:df:39 1500                                        Intel(R) Ethernet Controller X710 for 10GbE SFP+

vmnic8  0000:18:00.2 i40en       Up   10000Mbps  Full   3c:fd:fe:a7:df:3a 1500                                        Intel(R) Ethernet Controller X710 for 10GbE SFP+

vmnic9  0000:18:00.3 i40en       Up   10000Mbps  Full   3c:fd:fe:a7:df:3b 1500                                        Intel(R) Ethernet Controller X710 for 10GbE SFP+

[ivaylodyakov]

2017-10-12T06:08:11.588Z cpu24:72179)WARNING: Term: 1498: Unknown ANSI attribute 4

2017-10-12T06:08:11.588Z cpu24:72179)WARNING: Term: 1498: Unknown ANSI attribute 24

2017-10-12T06:08:11.589Z cpu24:72179)WARNING: Term: 1498: Unknown ANSI attribute 4

2017-10-12T06:08:11.589Z cpu24:72179)WARNING: Term: 1498: Unknown ANSI attribute 24

This is the last part of vmkernel log before the host is not responding.

[etniesBG]

The problem is solved with ESXi 6.0 u3 installation!

[mcs_che]

Did you try install again ESXI 6.5U1 and solve the problem or no?

I have similar problem on Fujitsu server:

Just stop responding two nic. Using it only for management access. After restart management interface on VMware console. It’s back to normal. Three identical servers. Three identical problem. Not in the same time. Sometimes one server per week, somethimes one server per monts. Custom Fujitsu image. I think it’s something bug with Ethernet drivers.