From Bugzilla Helper: User-Agent: Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.0; T312461; .NET CLR 1.0.3705; .NET CLR 1.1.4322) Description of problem: The bonding is configured on the ethernet interface eth2 and eth3. After rebooting, the MAC ADDRESS of the master interface is the one of eth0. The MAC address of eth0 is also become the one of eth2... Please could you help me?? My redhat release is RHEL ES 3.0 U2. Thanks Version-Release number of selected component (if applicable): initscripts-7.31.13.EL-1 How reproducible: Always Steps to Reproduce: 1. Configure eth2 and eth3 as slave interfaces 2. Reboot 3. Actual Results: MAC address of the MASTER interface is the one of eth0 Expected Results: MAC address of the MASTER interface is the one of eth2 or eth3 Additional info: My redhat release is RHEL ES 3.0 U2.
Please install a more recent package and try again. 7.31.18.EL is the latest.
*** Bug 143999 has been marked as a duplicate of this bug. ***
It seems that by adding the MAC address in the file of configuration of the slave interfaces (HWADDR) corresponding to the interfaces; it works fine. Do you know if do that is correct?
If your modules are loaded in a different order than the interfaces (or different than the probe order of the driver), yes, you need HWADDR to correctly assign the device name.
Is there a problem to put the initial MAC Address in the configuration files of the slave interface whereas the slave interfaces MAC address is normally the same after the bonding? (There is also a different MAC address in the configuration file of the slave interface and the command ifconfig -a for one slave interface)
HWADDR is a property of the hardware device used to identify which card is a slave before setting up the bonding device; it would be more-or-less expected for it to be different than the address that ends up on the bonding device.
Closing that it works as expected, based on comment #3.