3c905C hangings

[Casey Anderson]

I'm resending this since I didn't get a response before, and I now have some
new data.  See the original message for background.

I have a script that is watching eth0 in /proc/net/dev, and before the
computer hung (as far as connecting to it from the internet) the entries
looked something like:
Inter-|   Receive                                                |  Transmit
 face |bytes    packets errs drop fifo frame compressed multicast|bytes
packets errs drop fifo colls carrier compressed
  eth0:1579307866 91344950    0    0    1     0          0         0
2440096167 109854604    0    0    0 1559583       0          0
  eth0:1598049086 91637307    0    0    1     0          0         0
2488265585 110205755    0    0    0 1563232       0          0
  eth0:1615556514 91910791    0    0    1     0          0         0
2534570362 110542151    0    0    0 1566707       0          0
These entries are 5 minutes apart, read by a cron script.

Once my server hung, the entries were like the ones below, with the receive
fifo very large (relative to 1).
Inter-|   Receive                                                |  Transmit
 face |bytes    packets errs drop fifo frame compressed multicast|bytes
packets errs drop fifo colls carrier compressed
  eth0:1616550700 91926269    0    0 18756     0          0         0
2540153172 110567926    0    0    0 1566832       0          0
  eth0:1616550700 91926269    0    0 24866     0          0         0
2540168586 110568293    0    0    0 1566832       0          0
  eth0:1616550700 91926269    0    0 29048     0          0         0
2540182908 110568634    0    0    0 1566832       0          0

Does this help?  I'm using 2.2.12 kernel, with one machine SMP and the other
non-SMP.  Both are dual PIII-600 computers with 384MB ram.  They are game
servers so consequently there are very network intensive.  I'm using the
0.99H driver.  Also, I have a lot of "eth0: Transmit error, Tx status
register 82." messages.

I'm looking for a solution to the hanging. :)  I'm willing to take any
suggestions or questions on how to debug this more, suggestions how possible
fixes, or even an explanation of why it's happening (even if you don't have
a solution).  I'm grasping at straws here and really could use some help.  I
had the non-SMP machine go down twice within 24 hours and am now at my wits
end.  HELP!!!

Thanks,
Casey Anderson
casey.anderson@sierra.com


-----Original Message-----
From: Casey Anderson [mailto:Casey.Anderson@sierra.com]
Sent: Monday, March 06, 2000 11:36 AM
To: 'linux-vortex-bug@beowulf.gsfc.nasa.gov'
Subject: 3c905C hangings


I have several redhat 6.1 servers with new 3c905C network cards in them and
I keep getting what appears to be random hangs.  The are at a remote site,
but the one time I had someone ifconfig down; ifconfig up, the system
started responding to network traffic again.  I am currently using the 0.99H
driver (default with RH 6.1).  I am also getting lots of "eth0: Transmit
error, Tx status register 82." messages.  If anyone has any suggestions on
how to fix this, or at least how to help diagnose this I would be very
grateful.

Thanks in advance,
Casey Anderson
casey.anderson@sierra.com
-------------------------------------------------------------------
To unsubscribe send a message body containing "unsubscribe"
to linux-vortex-bug-request@beowulf.org
-------------------------------------------------------------------
To unsubscribe send a message body containing "unsubscribe"
to linux-vortex-bug-request@beowulf.org