Transmit timeouts with 1.06/1.09j

[Kerrin Pine]

Hi all,

Please correct me if this is the wrong place for this -- I have just built
a new server using an Asus P2B-LS m/b with onboard Intel 10/100 Ethernet..
we have had no trouble with these before, but I can't say we've ever put a
server under this load. It's a squid proxy server processing around
3-400kb/sec, connected to an Alteon L4 switch (CacheDirector).

The problem is that under heavy load, Ethernet stops working. These
messages are displayed on the console and also in /var/log/messages:

Aug  8 04:02:03 proxy2 kernel: eth0: Transmit timed out: status 0050  0090
at 17305561/17305576 command 000c0000.
Aug  8 04:02:03 proxy2 kernel: eth0: Trying to restart the transmitter...
Aug  8 04:02:08 proxy2 kernel: eth0: Transmit timed out: status 0050  0090
at 17305561/17305576 command 000c0000.
Aug  8 04:02:08 proxy2 kernel: eth0: Trying to restart the transmitter...
Aug  8 04:02:13 proxy2 kernel: eth0: Transmit timed out: status 0050  0090
at 17305561/17305576 command 000c0000.
Aug  8 04:02:13 proxy2 kernel: eth0: Trying to restart the transmitter...
Aug  8 04:02:18 proxy2 kernel: eth0: Transmit timed out: status 0050  0090
at 17305561/17305576 command 000c0000.
Aug  8 04:02:18 proxy2 kernel: eth0: Trying to restart the transmitter...
Aug  8 04:02:23 proxy2 kernel: eth0: Transmit timed out: status 0050  0090
at 17305561/17305576 command 000c0000.
Aug  8 04:02:23 proxy2 kernel: eth0: Trying to restart the transmitter...
Aug  8 04:02:28 proxy2 kernel: eth0: Transmit timed out: status 0050  0090
at 17305561/17305576 command 000c0000.

I am using Redhat 5.2 upgraded to kernel 2.2.10 -- the above is using
Donald's eepro100.c version 1.06 (the one that comes with 2.2.10 obviously).

I read about a few other similar problems.. and thought I would try the
latest test one for the hell of it.. went with version 1.09j.. this time
the server lasted a whole 15 seconds before giving up with this:

Aug  9 12:39:27 proxy2 kernel: eth0: Transmit timed out: status 0050  0070
at 365/377 command 000c0000.
Aug  9 12:39:27 proxy2 kernel: eth0: Trying to restart the transmitter...

Any ideas?

Some more diagnostic output is below..

Cheers
Kerrin

eth0: Intel EtherExpress Pro 10/100 at 0xa800, 00:E0:18:98:99:18, IRQ 10.
 Board assembly 668081-002, Physical connectors present: RJ45
 Primary interface chip i82555 PHY #1.
 General self-test: passed.
 Serial sub-system self-test: passed.
 Internal registers self-test: passed.
 ROM checksum self-test: passed (0x24c9f043).
 Receiver lock-up workaround activated.


eth0      Link encap:Ethernet  HWaddr 00:E0:18:98:99:18  
          inet addr:203.30.236.72  Bcast:203.30.236.127  Mask:255.255.255.192
          UP BROADCAST RUNNING MULTICAST  MTU:1500  Metric:1
          RX packets:1966743 errors:0 dropped:0 overruns:0 frame:0
          TX packets:2360674 errors:0 dropped:0 overruns:0 carrier:0
          collisions:0 txqueuelen:100 
          Interrupt:10 Base address:0xa800 

[root@proxy2 /root]# ./mii-diag 
Using the default interface 'eth0'.
Basic registers of MII PHY #1:  3000 782d 02a8 0150 05e1 4141 0001 ffff.
 Basic mode control register 0x3000: Auto-negotiation enabled.
 You have link beat, and everything is working OK.
 Your link partner can do 4141: 100baseTx-FD 10baseT-FD.

[root@proxy2 /root]# ./eepro100-diag  -m -f
eepro100-diag.c:v1.01 7/8/99 Donald Becker (becker@cesdis.gsfc.nasa.gov)
Index #1: Found a Intel 82557 EtherExpressPro100B adapter at 0xa800.
 MII PHY #1 transceiver registers:
  3000 782d 02a8 0150 05e1 4141 0001 ffff
  ffff ffff ffff ffff ffff ffff ffff ffff
  0a03 0000 0001 0000 0000 0000 0000 0000
  0000 0000 c205 0000 ffff ffff ffff ffff.


--
Kerrin Pine (kerrinp@chariot.net.au)   |  Web: http://www.chariot.net.au/
Operations Manager, Chariot Internet   |  Web: http://www.kerrin.com/
214 Greenhill Road, EASTWOOD SA 5063   |  Ph : (08) 8272 7677
PO Box 1046, UNLEY SA 5061             |  Fax: (08) 8272 7210

"I stayed up all night playing poker with tarot cards. I got
a full house and four people died." -- Steven Wright