VMkernel log fills with message [bnx2x_extract_max_cfg:1846(vmnicx)]Illegal...

[JPReyes]

Hi,

This is situation...

Upgrade a server Blade HP BL460G1 with vSphere 4.0 Update 2 to vSphere 4.0 Update 4, also installed two patches (ESX400-201112401 and ESX400-201203001).

------Bulletin ID------ -----Installed----- ----------------Summary----------------
ESX400-Update04         2012-04-03T11:00:58 VMware ESX 4.0 Complete Update 4
BCM-bnx2x-1.64.32.v40.1 2012-04-03T11:14:15 bnx2x: net driver for VMware ESX
ESX400-201203401-SG     2012-04-10T14:53:54 Updates VMkernel,VMX,hostd,VMware Tools
ESX400-201203402-SG     2012-04-10T14:53:54 Updates Python package
ESX400-201203403-SG     2012-04-10T14:53:54 Updates Curl RPM
ESX400-201203404-SG     2012-04-10T14:53:54 Updates samba RPM and libsmbclient
ESX400-201203405-SG     2012-04-10T14:53:54 Updates popt, rpm, rpm-libs, rpm-python
ESX400-201203406-SG     2012-04-10T14:53:54 Updates libuser
ESX400-201203407-SG     2012-04-10T14:53:54 Updates Kerberos RPMs
ESX400-201203408-BG     2012-04-10T14:53:54 Updates tzdata

This blade is in a enclosure C7000 with four Virtual Connect modules.

Review the logs files after of the update, found in the vmkernel log as follows:

Apr 10 15:15:46 host1 vmkernel: 0:00:17:04.399 cpu4:4207)<3>[bnx2x_extract_max_cfg:1846(vmnic7)]Illegal configuration detected for Max BW - using 100 instead
Apr 10 15:15:46 host1 vmkernel: 0:00:17:04.900 cpu2:4218)<3>[bnx2x_extract_max_cfg:1846(vmnic7)]Illegal configuration detected for Max BW - using 100 instead
Apr 10 15:15:47 host1 vmkernel: 0:00:17:05.402 cpu7:4219)<3>[bnx2x_extract_max_cfg:1846(vmnic7)]Illegal configuration detected for Max BW - using 100 instead
Apr 10 15:15:47 host1 vmkernel: 0:00:17:05.904 cpu5:4214)<3>[bnx2x_extract_max_cfg:1846(vmnic7)]Illegal configuration detected for Max BW - using 100 instead
Apr 10 15:15:48 host1 vmkernel: 0:00:17:06.406 cpu4:4205)<3>[bnx2x_extract_max_cfg:1846(vmnic7)]Illegal configuration detected for Max BW - using 100 instead

This message fills the log, not allowing other events can be displayed.

The vmnic7 not link because in the virtual connect profile no network assigned a this interface.

The problem with the message is solved when I enable and disable in the virtual connect profile of the blade any network a this interface...

Apr 10 15:16:05 host1 vmkernel: <6>bnx2x 0000:13:00.1: vmnic3: NIC Link is Up, 2500 Mbps full duplex, receive & transmit flow control ON
Apr 10 15:16:05 host1 vmkernel: 0:00:17:23.626 cpu4:4220)<3>[bnx2x_extract_max_cfg:1846(vmnic7)]Illegal configuration detected for Max BW - using 100 inste
ad
Apr 10 15:16:05 host1 vmkernel: 0:00:17:23.960 cpu7:4209)<3>[bnx2x_extract_max_cfg:1846(vmnic7)]Illegal configuration detected for Max BW - using 100 inste
ad
Apr 10 15:16:06 host1 vmkernel: 0:00:17:24.460 cpu1:4217)<3>[bnx2x_extract_max_cfg:1846(vmnic7)]Illegal configuration detected for Max BW - using 100 inste
ad
Apr 10 15:16:06 host1 vmkernel: <6>bnx2x 0000:13:00.5: vmnic7: NIC Link is Up, 2500 Mbps full duplex, receive & transmit flow control ON
Apr 10 15:20:09 host1 vmkernel: 0:00:21:27.998 cpu4:4208)<3>bnx2x 0000:13:00.5: vmnic7: NIC Link is Down
Apr 10 15:20:12 host1 vmkernel: <6>bnx2x 0000:13:00.7: vmnic9: NIC Link is Up, 3000 Mbps full duplex, receive & transmit flow control ON
Apr 10 15:20:13 host1 vmkernel: <6>bnx2x 0000:13:00.3: vmnic5: NIC Link is Up, 3000 Mbps full duplex, receive & transmit flow control ON
Apr 10 15:20:14 host1 vmkernel: <6>bnx2x 0000:13:00.1: vmnic3: NIC Link is Up, 3000 Mbps full duplex, receive & transmit flow control ON

However, when the server is reboot, the problem with message appears again.

I have not found any solution to this problem by searching google or VMware Kb.

Please, if anyone has had this same problem, could share how to fix.

Thanks.

[SimonStrutt · Accepted solution]

Have experienced a similar problem with HP BL495c G6's using the bnx2x driver.  It appears to be a problem with the source RedHat driver that gets turned into an ESX driver, and whilst its only a cosmetic logging problem, it renders your ESX logs pretty much useless.

If you install the latest driver bundle it resolves the problem (eg from http://downloads.vmware.com/d/details/dt_esxi40_broadcom_bcm57xxx_16211v401/ZHcqYnR0JWRiZCpwcA for ESX 4.0).  Go through the HP Support site to make sure you get the right driver bundle for your server, I got to the link above from the support page for a BL495 G6 not the BL460 G1 you have, though I'm assuming updated drivers should be available for your server too.

[JPReyes]

Install the driver bundle version recommend for Simon and the message dissapeared of the vmkernel log.

In HP Support not appears any driver bundle recommend for the model Blade 460c G1 and vSphere 4.0.

However, I´m surprised why had previously installed the driver bundle VMware ESX/ESXi 4.0 Driver CD for  Broadcom NetXtreme II NetXtreme II BCM57710, BCM57711, BCM57711E,  BCM57712  Ethernet Controllers version 1.64.32.v40.1 (http://downloads.vmware.com/d/details/dt_esx40_broadcom_bcm577xx/ZHcqYnQlZHdidGQlcA==) that appears as the last update driver for the model Broadcom NetXtreme having the Blade 460c G1 in the VMware Support Site but not found this driver in the HP Support Site as recommended.

Uninstall this driver bundle version 1.64.32.v40.1, install the driver bundle version recommended for Simon and the problem disappeared.

Lesson learned today?, search drivers in the HP Support first.

Thank Simon for your help.