DellR710 server BCM5709 Nic on RHEL5.x and vmwareesxi4.1

The data center uses DellR710 servers with BCM5709 Nic. The default RHEL5.x system driver does not support BCM5709 Nic. If the NIC encounters a large traffic, it will be hungup. This problem involves ACPI power management. When the network adapter is working normally, ACPI will mistakenly think that it is idle and then give it to hungup. In RHEL5.x

The data center uses Dell R710 servers with BCM5709 Nic.

The default RHEL5.x system driver does not support BCM5709 Nic. When the NIC encounters heavy traffic, it will be hung up.

This problem involves ACPI power management. When the network adapter is working normally, ACPI will mistakenly think that it is idle and then give it to hung up.

On RHEL 5.x or Centos 5.x, there are two solutions:

1. Use the built-in kernel driver to modify Kernel Parameters and disable.CpI:

Vi/boot/grub. conf
Title CentOS (2.6.18-164.6.1.el5)
Root (hd0, 0)
Kernel/vmlinuz-2.6.18-164.6.1.el5 ro root =/dev/VolGroup00/LogVol00 rhgb quiet acpi = off noapic
IniTrD/initrd-2.6.18-164.6.1.el5.img

 

The parameter acpi = off noapic is to be added. This solution has been tested to ensure that the Network Card's traffic reaches more than 300 MB.

2. download the latest driver upgrade from the Dell website.

 

This problem occursLinuxI got a good solution, but today I received several alarm emails from the machine room, saying that I had a serious packet loss on the Dell R710 server with vmware esxi 4.1. When it was high, it was 50%. I went on to check it out, packet Loss exists. I checked the esxi log and didn't see any errors. I suspect it is also a problem with the NIC Driver of BCM5709.

 

Then I went to www.vmware.com and found that the latest version is vmware esxi 4.1 up.Date1, carefully check the Release Notes (https://www.vmware.com/support/vsphere4/doc/vsp_esxi41_u1_rel_notes.html), which found the BCM5709 NIC Driver updates:

ESXi hosts might not boot or cause some devices not to be accessible when usingMoreThan 6 bnx2 ports
An error message similar to the following is displayEdIn/var/log/messages of ESXi: CPU10: 4118-intr vector: 290: out of interrupt vectors. Before applying this fix, bnx2 devices in MSI-X mode and jumbo frame configurationSuPport only 6 ports. The issue is resolved in this release. In this release, the bnx2 driver alloCatEs only 1 RX queue in MSI-X mode, supports 16 ports, and saves memory resources.

 

When the NIC Driver is updated, you are ready to upgrade to vmware 4.1 update 1, find a machine to install vmware-vsphere-cli, and start the upgrade:

1. Download the upgrade package from vmware: update-from-esxi4.1-4.1_update01.Zip

2. Disable all running VMS

3. Set esxi 4.1 to maintenance mode

4. Upgrade vmware-vsphere-cli on the machine that has just Installed vmware-vsphere-cli:

Vihostupdate -- server x. x -- username root -- passwordPasswd-I-B update-from-esxi4.1-4.1_update01.zip-B ESXi410-Update01
After running:
Please waitPatchInstallation is in progress...
The update completed successfully, but the system needs to be rebooted for the changes to be updated Ti
Ve.

5. reboot esxi

 

After the restart, ssh is connected to the esxi console:

~ # Ethtool-I vmnic0
Driver: bnx2
Version: 2.0.7d-3vmw
FiRmWare-version: 5.0.13 bc 5.0.11 NCSI 2.0.5
Bus-info:. 0

What we can see before the upgrade is:

~ # Ethtool-I vmnic0
Driver: bnx2
Version: 2.0.7d-2vmw
Firmware-version: 5.0.13 bc 5.0.11 NCSI 2.0.5
Bus-info:. 0

Not to force it, the main version is the same, worried about the problem, continue to find the driver, found that vmware really have: http://downloads.vmware.com/d/info/datacenter_downloads/vmware_vsphere_hypervisor_esxi/4_0#drivers_tools

VMware ESX/ESXi 4.1 Driver CD for Broadcom NetXtreme II Ethernet Network Controllers

VMware has made an ISO for Broadcom NetXtreme II Ethernet Network Controllers. It seems that this Nic has many driver problems. Download this ISO at the address below:

Http://downloads.vmware.com/d/details/esx41_broadcom_netextremeii_dt/ZCV0YnRlaHRidHdw

After downloadExtractGo to bcm-bnx2-2.0.15g.8.v41.1-offline_bundle-325733.zip. Upgrade:

Vihostupdate -- server x. x -- username root -- password passwd-I-B BCM-bnx2-2.0.15g.8.v41.1-offline_bundle-325733.zip

After installation, reboot esxi,

Let's look at the driver version:

~ # Ethtool-I vmnic0
Driver: bnx2
Version: 2.0.15g.8.v41.1
Firmware-version: 5.0.13 bc 5.0.11 NCSI 2.0.5
Bus-info:. 0

This is much better, haha.

Then exit maintenance mode and start the virtual machine.

I don't know how the problem works. I can only observe whether the problem can be solved.