In my super count cluster, there is a server failure repair, the motherboard is broken, replace the motherboard, unable to identify the network card, with the command ifconfig-a to see only the LO loopback 127.0.0.1, and eth7,eth8,eth9 and so on no network card information, so, Causes server and cluster node servers to not ping
Causes and solutions:
The system load NIC driver will read a file, this file is a buffer file, including the MAC address of the network card, because the replacement of the motherboard, the MAC address of the network card has changed, but the MAC address of the file has not changed, or the previous bad motherboard of the above network card MAC address, so that the system in the Load network card, Read this file when reading is the MAC address of the previous network card, and now replace the motherboard after the NIC MAC address inconsistency caused confusion, so the current network card is not recognized;
This file is:/etc/udev/rules.d/70-persistent-net.rules, after the deletion of this file (or rename the file) to restart the server can be resolved, the restart will regenerate the file;
Note: Sometimes after restarting the above situation, compared to the previous file, the new file MAC address and name has changed, Name=eth0, and the name of the previous name=em1 is different;
Action: ifconfig eth0 down
Then edit: Vi/etc/sysconfig/network-scripts/ifcfg-eth0
Restart Network card: Service Network restart
Linux server does not recognize the network card after replacing the motherboard solution