VMware Cloud Community
NabhanWW
Contributor
Contributor

Server suddenly not reachable. Problem with NIC?

Hey there,

we're running a VMware ESXi 5.1.0 build-1312874 with a Realtek Realtek 8168 Gigabit Ethernet NIC. The Server ran about 265 days without any problems. But last night, the server wasn't reachable for about 20mins. As I said, the server doesnt seem to be crashed or someting, since the uptime command says something different. I didn't see any errors in the error log except the following:

2014-08-04T23:31:32.403Z cpu0:4554)WARNING: LinNet: netdev_watchdog:3220:NETDEV WATCHDOG: vmnic0: transmit timed out

Has anyone already experienced such problem? If you need further information, let me know.

Thanks in advance

Best regards

-Norman

Tags (2)
Reply
0 Kudos
8 Replies
SG1234
Enthusiast
Enthusiast

Reply
0 Kudos
NabhanWW
Contributor
Contributor

Hi,

thanks for your hint. But as far as I understand this KB article, only ESX/ESXi 4.x with Broadcom NICs are affected by this? Moreover it has been resolved.

Best regards

Reply
0 Kudos
SG1234
Enthusiast
Enthusiast

version is showing 5.1 as well - can you confirm it is bnx2 driver in your server

Reply
0 Kudos
NabhanWW
Contributor
Contributor

The driver is r8168. Its a Realtek NIC, where did you see the link to Broadcom NICs? Here are all infos from the nic:

~ # esxcli network nic list

Name    PCI Device     Driver  Link  Speed  Duplex  MAC Address         MTU  Description

------  -------------  ------  ----  -----  ------  -----------------  ----  -------------------------------------

vmnic0  0000:003:00.0  r8168   Up     1000  Full    d4:3d:7e:ed:0a:cd  1500  Realtek Realtek 8168 Gigabit Ethernet

~ # ethtool -i vmnic0

driver: r8168

version: 8.013.00-NAPI

firmware-version:

bus-info: 0000:03:00.0

~ # vmkchdev -l |grep vmnic0

00:03:00.0 10ec:8168 1462:7816 vmkernel vmnic0

Reply
0 Kudos
NabhanWW
Contributor
Contributor

Update:

I think, can confirm that the bnx2 driver is installed on the server.

~ # esxcli software vib list | egrep bnx2

net-bnx2                      2.0.15g.v50.11-7vmw.510.1.12.1065491  VMware  VMwareCertified  2013-11-12

net-bnx2x                      1.61.15.v50.3-1vmw.510.0.11.1063671  VMware  VMwareCertified  2013-11-12

scsi-bnx2i                    1.9.1d.v50.1-5vmw.510.0.0.799733      VMware  VMwareCertified  2013-11-12

Is this what you meant? But i dont understand why its relevant, because we use another NIC. Or is this irrelevant, since its just built-in from stock?

Many thanks and best regards.

Reply
0 Kudos
SG1234
Enthusiast
Enthusiast

yeah - i may have misread  that its broadcom nic --  was the traffic too high at that time   - also would like to advise to check any logs on the physical switch port

HTH

Reply
0 Kudos
NabhanWW
Contributor
Contributor

We already asked our isp. Their systems hadn't recognized any suspicious actions. Any other ideas?

Best regards.

Reply
0 Kudos
RAJ_RAJ
Expert
Expert

Hi ,

you can delete the vmk0 from command and create new one and assign IP .

  1. List vmkernel interfaces: # esxcli network ip interface list
  2. Remove vmkernel interface(s): # esxcli network ip interface remove –i vmkX (where X is the vmk# you want to remove)
  3. List portgroups: # esxcli network vswitch standard portgroup list
  4. Remove Management Network portgroup: # esxcli network vswitch standard portgroup remove –p ‘Management Network’ –v vSwitchX (where X is the vSwitch number)
  5. Remove vmnics: # esxcli network vswitch standard uplink remove –u vmnicX –v vSwitchX (where X is the uplink number and vSwitch number)
  6. Add new Management Network portgroup: # esxcli network vswitch standard portgroup add –p ‘Management Network’ –v vSwitchX
  7. Add new vmkernel interface: # esxcli network ip interface add –i vmk0 –p ‘Management Network’ –m 1500
  8. Configure the new vmkernel interface: # esxcli network ip interface ipv4 set –i vmk0 –I <IP_address> -N <Network_Mask> -t static
  9. Add vmnics: # esxcli network vswitch standard uplink add –u vmic0 –v vSwitch0

Note-  -i , -p gives error you may have to type -p=‘Management Network’  .

If this server is running HP blade chassis , then MAC can be an issue .

Solution

Unassign profile from blade  (Don't delete the profile )

Create new profile and assign the Blade

power on server and check .

later you can rename and delete old profile.

RAJESH RADHAKRISHNAN VCA -DCV/WM/Cloud,VCP 5 - DCV/DT/CLOUD, ,VCP6-DCV, EMCISA,EMCSA,MCTS,MCPS,BCFA https://ae.linkedin.com/in/rajesh-radhakrishnan-76269335 Mark my post as "helpful" or "correct" if I've helped resolve or answered your query!
Reply
0 Kudos