John R Pierce wrote:
a previously rock solid reliable server of mine crashed last night, the server was still running but eth0, a Intel 82574L using the e1000e driver, went down. The server has a Supermicro X8DTE-F (dual Xeon X5650, yada yada). server is a drbd master, so that was the first thing to notice network issues. Just a couple days ago I ran yum update to the latest, I do this about once a month.
/var/log/messages logged...
<snip>
(prior to this was nothing but normal smbd complaining about CUPS not configured). Duplex, Flow Control: Rx/Tx May 9 22:52:56 sg1 kernel: e1000e: eth0 NIC Link is Down May 9 22:53:01 sg1 kernel: e1000e: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx May 9 22:55:30 sg1 kernel: e1000e: eth0 NIC Link is Down May 9 22:55:35 sg1 kernel: e1000e: eth0 NIC Link is Up 1000 Mbps Full Duplex, Flow Control: Rx/Tx
<snip> This is going to sound really stupid, but consider replacing the patch cord. If that doesn't work... should I assume that this m/b has at least two embedded NIC? If so, try using the other NIC.
mark "I haven't done both of those in the last few months, no...."