Updated from 4.0U2 to 4.1, network adapaters fail.

[CatHat]

Hello there!

Just 10 minutes ago i updated my esxi whitebox from 4.0U2 (working perfectly) to 4.1 through the vmware vSphere CLI. The update and everything worked perfectly until i tried to connect to the machine. if i do a ping to the machine from my windows host it replies until i try to connect to esxi with https/vSphere client/ssh, then it just time-outs for a while and i cant connect.

I checked the server "(1) messages log" and it says:

e1000_clean_tx_irq : Detected TX unit hang

netdev_watchdog : NETDEV WATCHDOG : vmnic2 : transmit timed out

and after a while:

vmnic2 is down. Affected portgrupp: Management network. 0 uplinks up. Failed criteria : 130

and after a while it comes back up:

Uplink vmnic2 had recovered from a transient failure due to watchdog timeout.

I cant export the log since i cant connect to the machine 😕

What should i do to fix this, is it a faulty e1000 driver shipped with the update?

My setup mb and nic:

Asus M4A785TD-V EVO, Socket-AM3 and Intel® PRO/1000 MT Dual Port Server Adapter

Tanks!

[DSTAVERT]

: it was at 100 mbit and setting it manually had no effect on the problem.

Are you connected to a 100 mbit switch? Try 1GB

[CatHat]

I connected my laptop directly to the esxi box, tried both auto (1000 mbit) and setting it manually, same problem......

Darn im really out of ideas.......

[CatHat]

One thing tho, if you check the HCL for my DID it says:

ESX / ESXi 4.1	e1000 version 8.0.3.2	inbox
ESX / ESXi 4.0 U2	e1000 version 8.0.3.1	inbox
ESX / ESXi 4.0 U1	e1000 version 8.0.3.1	inbox
ESX / ESXi 4.0	e1000 version 8.0.3.1	inbox

Would it be possible to use the e1000 version that works from some earlier release?

[DSTAVERT]

Sorry. Out of ideas.

If you have at least 8GB RAM and support for hardware virtualization you might want to look at VMware workstation. You can install ESXi as a guest and VMs inside that. You can install more than 1 ESXi guest and set up things like vmotion etc.

It isn't that hard to get server hardware that supports ESXi. Off lease, end of model, returns.

[DSTAVERT]

You can try copying the modules to the new install but you risk making an unstable host.

[CatHat]

Well thanks for all the effort you put in this anyway

I might consider changing to workstation/server but for now i think i will just continue running esxi 4.0U2. Doesn't it get allot slower if you run OS->workstation->esxi->OSes? (so many layers to the hardware?, but i got 8 gb ram and AMD-V support)

[DSTAVERT]

Download Workstation trial and try it. Without lots of RAM (more than 😎 it is always a tight squeeze running 2 ESXi hosts inside workstation but it does allow you to test out different scenarios that would otherwise require more than one host.

Good luck. Hope there is an update that fixes your issue.

[kmyerqsv]

I am experiencing a near identical problem as the original poster. I had a prototype server that had ESXi 4.0U1 and then U2 installed on it, running a few VMs. The original server had an Adaptec RAID card in it, but since there is no CIM support for Adaptec HBAs, and we'd really like to know about drive failures, we switched that out for a LSI RAID card. When I made that change, I also upgrade to ESXi 4.1, and immediately began to experience problems.

Sequence of events:

Install ESXi 4.1 (build 260247) to USB flash drive on server, using IPMI virtual CD - worked flawlessly

Attempt to install base VM (Centos 5.5, using netinstall) works fine until it starts to download the second stage image for install over the network; network stops responding, and eventually the server gives a PSOD

Thinking there may be an updated driver in a newer firmware release, attempt to update to this using the CLI interface on my PC; after the image starts transferring, the server NIC locks up, and again, eventually goes to a PSOD

Remove LSI RAID card from slot and install HP Broadcomm NIC; am able to update to the latest firmware without any hangs or trouble. However, that also means I lose persistent storage, so I can't provision VMs with the Broadcomm NIC installed. Remove Broadcomm NIC and boot on latest ESXi version (build 320137).

Attempt to create new VM; after the install image starts to transfer over the network, the following is logged:

e1000_clean_tx_irq : Detected TX unit hang

Then, the server gives a PSOD (so same behavior as the original 4.1 install).

The hardware is partly white box but has components which do show up on the HCL.

Motherboard is a Supermicro X8-SILF, with a Xeon 3430, and 4Gb of RAM (2 DIMMs of X8 sticks @1333MHz)

Dual onboad Intel NICs (listed as Intel Corporation 82574L Gigabit Network Connection, PCI id is 8086:10d3 15d9:0605)

LSI SAS3041E-R RAID card

I'd consider using the Broadcomm NIC that works, except I have a single PCI-e slot to use (with a riser card), since the case is a 1U server case, and that slot is tied up with the RAID card.

Any suggestions on what else to try?

Thanks!

[DSTAVERT]

Create a new post and copy your content to there. You will get better responses in a fresh post. You can refer to this thread.

Forum Upgrade Notice - We will be upgrading VMware Communities systems between 10-12 December 2010. During this time, the system will be placed in READ-ONLY mode.