VMware Cloud Community
AlexNG_
Enthusiast
Enthusiast

Network devices disappears

Hi all,

I'm deploying some ESX in a client and, swomething very weird happened, let me explain:

The ESX server taht has a weird behaviour is an IBM x3550, full of CPUs and RAM, 2 hbas, 2 onboard nics + a quad port Intel eth adapter.

Two days ago, the ESX had the two onboard nics for SC and VMotion (with VLAN tagging). The rest, were in vSwitch1 and providing service to 10 portgroups (each one in a VLAN). Ok, so everithing was up and runinng!

Today, our customer called me saying that VMs on that server had no network link. Our customer has still no access to remove nics from vSwitch nor things like that. Once VMs have been vmotioned and host in maintenance mode, I've seen that the vSwitch1 (4 eth devices, 10 VLANS, and everithing ok...) has no vmnics assigned!!!! What I checked after that is that the links we're up to reassign them to vswitch1 but, here is the surprise and weird thing:

no additional vmnics are present in the host! WOW!!! They've disapeared!!!!! I've checked phisycally the server and the quad port adapter is showing network activity!!

I've did a lspci, and it just shows vmnic0 and vmnic1, vmnics 2, 3, 4 and 5 are no longer present!

inside vmkwarning, I've seen that 12 May at 13:10:01 I've the following message:

vmkernel: 0:00:00:02.979 cpu3:1036)WARNING: PCI: 1967: No such device

I've checked vmkernel logs, hostd logs and everithing seems ok. On the boot.log (last boot, a day before), the server lists all de vmnics correctly.

The eth adapter is a Intel PRO/1000 PT Low Profile Quad Port PCIe Gigabit Ethernet Contoller

Actually I cannot open an SR due to external non-technical factors...

As anyone had such issue? Any hint clue?

Many thanks in advance!

Alex

Message was edited by: AlexNG_

Message was edited by: Texiwill fixed Subject to not show , only

If you find this information useful, please award points for "correct" / "helpful".
0 Kudos
6 Replies
AlexNG_
Enthusiast
Enthusiast

Update.

The boot log that I checked wasn't last boot log. On the last boot log there are messages like:

esxcfg-vswitch: Unable to restore uplink vmnic2 of virtual swtich vSwitch1. Reason was Not a valid pnic: vmnic2: . Not a valid pnic: vmnic2

So seems like the server, physically, in the last boot has not seen the pci card... lloks like hardware problem!

If you find this information useful, please award points for "correct" / "helpful".
0 Kudos
depping
Leadership
Leadership

Sounds like a hardware issue indeed.

Duncan

VMware Communities User Moderator | VCP | VCDX

-


Blogging:

Twitter:

If you find this information useful, please award points for "correct" or "helpful".

AlexNG_
Enthusiast
Enthusiast

Hi depping,

yes right I'm near 100% sure you're right. I've found that article at IBM:

http://www-947.ibm.com/systems/support/supportsite.wss/docdisplay?lndocid=MIGR-5079623&brandind=5000...

As many other times, the symptons are very very very similar to what I'm seeing here, and the affected configuration is also veeeeeeeery close also. I'm opening an SR with IBM to check it and/or fix it.

Will update with the results.

Many thanks!!

Alex.

If you find this information useful, please award points for "correct" / "helpful".
0 Kudos
coco26
Contributor
Contributor

any resolution on this? were the cards dead or driver issue possibly?

0 Kudos
AlexNG_
Enthusiast
Enthusiast

Hi coco,

Dooooh!!! Sorry for the delay!!! As you can read in my last post, I was unsure if the IBM technote would apply to our case. Afetr opening a SR with IBM, they confirmed that there were a hard issue and it should be replaced!

After the card change, everiting up and runing as expected!!

AlexNG

If you find this information useful, please award points for "correct" / "helpful".
0 Kudos
coconchuir
Contributor
Contributor

great thanks, suspected as much

0 Kudos