On a recent install of RH 6.1 on a single processor(dual capable) intel cserver 210, the network hangs when cpu usage is at 0% and there is still plenty of memory available. When pinging the server I get latencies around 28000+ ms and can almost not log into the server through telnet. I can only send about 10 cpm in a telnet session. I have several other servers on the same network that I can connect to just fine. This server is running sendmail-8.9.3-11mdk and apache-1.3.9-7mdk.
Does "ifconfig -a" show collisions and/or errors? You might also look at "netstat -s" for unusual behavior.
hmm... I have the same problem.... I'm running on a single P2 366 mhz, ide & scsi disk's, 128 mb ram, and an Intel Etherexpress NIC i get the following in /var/log/messages Dec 16 10:30:44 wroom kernel: eth0: Transmit timed out: status 0050 0070 at 16683/16 695 command 000c0000. Dec 16 10:30:44 wroom kernel: eth0: Trying to restart the transmitter... Dec 16 10:31:14 wroom kernel: eth0: Transmit timed out: status 0050 0070 at 29222/29 234 command 000c0000. Dec 16 10:31:14 wroom kernel: eth0: Trying to restart the transmitter... Dec 16 10:31:42 wroom kernel: eth0: Transmit timed out: status 0050 0070 at 40307/40 319 command 000c0000. Dec 16 10:31:42 wroom kernel: eth0: Trying to restart the transmitter...
I am having a similar problem. While copying files via a samba share. TCP/IP seems to "go away" The only way to get the shares started again is to stop and start the network (not samba). I'm running on a P133 with dual Intel Pro/10 Ethernet cards. I am getting a large number of collisions on the internal LAN ethernet card (which has all my PCs connected to it. Any ideas?
There are people workjing on eepro10 problems yes
This should be fixed in the 2.4 kernel included in Red Hat Linux 7.3 which has an updated eepro100 driver, as well as the alternative e100 driver from intel. Please reopen if the problem persists in new releases.