CentOS 6.4 Frame errors and dropped packets

[rupesh]

Hello All,

I am seeing a lot of dropped packets and frame errors on 5 nodes, running an application and all the nodes connected to a single switch. I am using a 4 port network adapter (Ethernet controller: Broadcom Corporation NetXtreme BCM5719 Gigabit Ethernet PCIe) on HP DL380p Gen8's. Out of 4 ports, I am only using 1 port on each server. I have updated the firmware for the network card and so is running the latest version. I have replaced the Ethernet cables for all the 5 nodes and also swapped the switch with a different model/vendor, moved the cable to a different port on the network card on every node and did the same test on the switch side too. There are no other nodes connected to the switch except for these 5 nodes. Also played with MTU sizes and jumbo frames. But none of these seems to fix the issue, and still see a number of [b]dropped packets and frames.[/b]

Did any one experience this kind of issues? And any suggestions here? And if there are any network tools that I can try to test?

I have opened cases with the hardware vendors for both the switch and the servers. They do not find any hardware related issues on either of them. So, I am now concerned about the [b]driver being used for this card?[/b] Below are the specs:

Lspci -v:

Ethernet controller: Broadcom Corporation NetXtreme BCM5719 Gigabit Ethernet PCIe (rev 01)
Subsystem: Hewlett-Packard Company Device 169d
Flags: bus master, fast devsel, latency 0, IRQ 32
Memory at f6bf0000 (64-bit, prefetchable) [size=64K]
Memory at f6be0000 (64-bit, prefetchable) [size=64K]
Memory at f6bd0000 (64-bit, prefetchable) [size=64K]
[virtual] Expansion ROM at f6b00000 [disabled] [size=256K]
Capabilities: [48] Power Management version 3
Capabilities: [50] Vital Product Data
Capabilities: [58] MSI: Enable- Count=1/8 Maskable- 64bit+
Capabilities: [a0] MSI-X: Enable+ Count=17 Masked-
Capabilities: [ac] Express Endpoint, MSI 00
Capabilities: [100] Advanced Error Reporting
Capabilities: [13c] Device Serial Number 00-00-2c-76-8a-56-31-d0
Capabilities: [150] Power Budgeting
Capabilities: [160] Virtual Channel
Kernel driver in use: tg3
Kernel modules: tg3


modinfo tg3 :

filename: /lib/modules/2.6.32-358.el6.x86_64/extra/tg3/tg3.ko
firmware: tigon/tg3_tso5.bin
firmware: tigon/tg3_tso.bin
firmware: tigon/tg3.bin
version: 3.132f
license: GPL
description: Broadcom Tigon3 ethernet driver


Uname -a:

Linux example.example.com 2.6.32-358.el6.x86_64 #1 SMP Fri Feb 22 00:31:26 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux

Ifconfig:

Link encap:Ethernet HWaddr 2C:76:8A:56:31:D0
inet addr:10.49.61.111 Bcast:10.49.61.255 Mask:255.255.255.0
inet6 addr: fe80::2e76:8aff:fe56:31d0/64 Scope:Link
UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1
RX packets:19389481 errors:0 dropped:312 overruns:0 frame:72
TX packets:19141824 errors:0 dropped:0 overruns:0 carrier:0
collisions:0 txqueuelen:1000
RX bytes:4225077789 (3.9 GiB) TX bytes:3883973433 (3.6 GiB)
Interrupt:32



Any help or suggestions are truly appreciated!!


Thank You!

Regards,
Rupesh

[TrevorH]

Your first step should be to run `yum update` and reboot into the latest kernel.

[Modeler]

Hi,

Sorry to bump an old thread but I have exactly the same issue. Did a kernel update resolve this as TrevorH suggested?

Thanks.

[TrevorH]

No-one should run a 6.4 kernel in any case. There have been numerous security vulnerabilities discovered and fixed since then.