after upgrade 5.5

After upgrade to 5.5, the system seems to have dropped off the face of the earth. no ssh nothing on port 10000. if i could not use private IP to ssh i would be up a creek.

Any other reports?

Status: 
Closed (fixed)

Comments

I haven't seen any other reports like this, and Cloudmin shouldn't ever touch the network configuration on the master system anyway.

Which variant are you running? GPL, full or physical systems?

Full. The only change was 5.5 and now there is no routing at all any ideas on where to look?

What does netstat -rn output on your system?

Also, what does /etc/sysconfig/network-scripts/ifcfg-eth0 contain? It sounds like it may be missing a GATEWAY line..

netstat produces all private network

eth0 does not contain a gateway line it is in /etc/sysconfig/network

I have three Identical dell servers all running 5.6 x64 and the now latest kernel 2.6.18-238.9.1.el5xen. so the patch level is the same. The server that runs cloudmin is the only one that has suddenly dropped all outside networking. I see the eth0 interface come up it pings for about 30 seconds and then drops off the face of the earth. Nothing is logged anywhere that I have found. I even have xenbr0 and vif0.x; just no peth0 and no eth0; (if memory servers now that i have been booting in and out of the xen kernel oddly enough ifconfig -a shows all interfaces WOW.

Seems that it has lost its bridge or configuration but at this point I am at a bit of a loss.

Any ideas or further direction would be appreciated.

You might want to try disabling all Xen instances at boot time, just in case one of them is breaking networking somehow when it comes up.

Also, are you running the Xen kernel and tools that come with CentOS 5.6, or packages from another source?

disabled all vms all standard kernel and sources Thankfully "xen" instances on this server were in house and non production.

Could not find anything log wise etc purely coincidental that the network dropped off the face of the earth within hours of my updating to CM 5.5. Finally found a single odd log entry:

bnx2: fw sync timeout, reset code = 1030009 which leads me to believe there is some ODD MSI issue with the card that when the bridge is created it just fails and does not log the error.

Finally found a difference in Firmware version on the nics on the troubled machine vs the working xen servers updated the nic firmware and all is happy. Gotta love what we do :P

Please close and Thanks for the help\direction\moral support :)

A network card issue certainly sounds like a possible cause .. I will close this ticket then.