Let me just state up front that I’m not a huge fan of VMWare products these days. Sure, the technology is impressive and it’s very useful - but I’ve always run into odd quirks with their software. I suppose I like their desktop products better than their server products, though. Perhaps that’s because I had such a good experience with Microsoft Virtual Server and my ability to easily set up “virtual labs” through scripting…

In any case, I ran into an odd problem last night after some network downtime. I brought the network back up, but the VM’s still weren’t communicating over the second ethernet adapter, configured for bridging. Grepping through dmesg, I noticed the following errors:

e1000: eth1: e1000_watchdog: NIC Link is Up 1000 Mbps Full Duplex e1000: eth1: e1000_watchdog: NIC Link is Down e1000: eth1: e1000_watchdog: NIC Link is Up 1000 Mbps Full Duplex e1000: eth1: e1000_watchdog: NIC Link is Down e1000: eth1: e1000_reset: Hardware Error

…and then there were no more messages about eth1. Googling around a little bit indicated a possible solution would be reloading the kernel module, so…

rmmod e1000; modprobe e1000 ifconfig eth1 up

And then traffic started flowing again!

bridge-eth1: enabling the bridge device eth1 entered promiscuous mode audit(1191399073.815:34): dev=eth1 prom=256 old_prom=0 auid=4294967295 bridge-eth1: enabled promiscuous mode bridge-eth1: up e1000: eth1: e1000_watchdog: NIC Link is Up 1000 Mbps Full Duplex

Note that the “ifconfig eth1 up” is essential for the vmware bridge to realize the interface came back and it can start bridging info over it again. Unfortunately, it took me several minutes and the realization that I still wasn’t getting link lights after loading the module to realize this…

That’s my tech tip of the day, back to reporting!