Larry Finger wrote:

On 03/18/2013 02:08 PM, Per Jessen wrote:

...

See https://bugzilla.novell.com/show_bug.cgi?id=809843

That this wasn't picked up until after GM shows how little testing we really get done. My own tests on dual-nic boxes did not reveal this issue because the driver always ordered the devices correctly.

Testing coverage will always be a problem. Many of us start out by installing in a VM, thus the hardware is very limited. I usually switch to real machines at MS4 or 5, but my systems are pretty vanilla. That goes for both hardware and software.

I have access to plenty and varied hardware, but I have to match my testing to whatever is happening work-wise. During most of 12.3 development, I was working with iSCSI on some plain supermicro boxes with dual intel nics, nothing very unusual. I wish we had a system where cases like this would/could create a new test-case - "Test naming of multiple NICs" or some such. It wouldn't have to be overly elaborate, but if testers could just increment the counter, then towards the GM, we might have an idea of how much testing had actually been done. For iSCSI, I think I'm pretty much the _only_ tester, it really to be labelled "experimental" :-(

One additional factor is that anyone who upgraded from 12.2 to 12.3 still has a 70-persistent-net.rule; at least my system does.

The 12.3 system curiously also have the file, but it's empty. I don't really understand the problem in simply keeping the persistent-net rules. /Per -- Per Jessen, Zürich (5.4°C) http://www.dns24.ch/ - free DNS hosting, made in Switzerland. -- To unsubscribe, e-mail: opensuse-testing+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-testing+owner@opensuse.org