Which GigE NIC for reliable use?

From: Steve Kargl <sgk_at_troutmask.apl.washington.edu>
Date: Thu, 21 Jun 2007 09:07:43 -0700
I've been experiencing problems with the bge device 
for several weeks.  In this time, I've tried tuning
every imaginable parameter that I could find.

There appear to be several related problems:

node10:kargl[203] netstat -I bge1
Name    Mtu Network       Address              Ipkts Ierrs    Opkts Oerrs  Coll
bge1   9000 <Link#2>      00:e0:81:40:48:93 81505160 238721 81933513     9     0
bge1   9000 192.168.0.0   node10            81504878     - 81933689     -     -

Notice the Ierrs value continuously grows with the MPI application
I have runs.

In /var/log/messages one finds:

Jun 20 23:20:42 node10 kernel: bge1: watchdog timeout -- resetting
Jun 20 23:20:42 node10 kernel: bge1: link state changed to DOWN
Jun 20 23:20:46 node10 kernel: bge1: link state changed to UP

This DOWN/UP breaks the MPI application and leads to several
additional messeages of the form. 

Jun 20 23:22:33 node10 kernel: TCP: [10.208.78.111]:54801 to 
[10.208.78.111]:49376 tcpflags 0x10<ACK>; syncache_expand: Segment failed
SYNCOOKIE authentication, segment rejected (probably spoofed)

So, I plan to replace all of the bge devices with a reliable,
robust GigE NIC.  Anyone have a suggestion for such a cards?

--
Steve
Received on Thu Jun 21 2007 - 14:09:07 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:13 UTC