What are the drawbacks of using consistent network naming ?

asked 2019-09-12 00:07:44 -0500

updated 2019-09-12 05:10:32 -0500

Hi All,

We would like to know what are the drawbacks of using consistent naming ?. We enabled consistent naming and now Ethernet interfaces are named in the order of pci bus. Now we would like to know will this have an impact when we do live migration ?.

Btw if yes, how can a udev rule help here ?. We created a udev rule for persistent naming and this depends on lspci as we want interface in the bus order.

Can we assume lspci bus will remain the same across migration ? looks naive but I think can't. Then how do we handle such case ?. Any pointers on the same would be helpful.

https://bugs.launchpad.net/mos/+bug/1501430 (link text)

As per above link looks like its not necessary that it will be given in the same order as in libvirt.xml, can someone clarify on the same ?.

rgds Balaji Kamal Kannadassan

edit retag flag offensive close merge delete