We indeed should keep the module naming consistent, but I think it has 2 layers meaning, the one is module naming should be consistent; the other is related module should depict similar function, for example,
kvm_monitor.py: Interfaces to the QEMU monitor kvm_vm.py: Utility classes and functions to handle Virtual Machine creation using qemu. For libvirt-autotest: libvirt_monitor.py: (we haven't any definition in here) libvirt_vm.py: Utility classes and functions to handle Virtual Machine creation using libvirt. For libvirt, you know it hasn't a monitor concept like qemu, virsh has only a qemu-monitor-command command to directly send qemu command to guest bypass libvirt, if we put all of virsh commands into libvirt_monitor module, I don't know how to define the module function. In addition, libvirt has also some host related virsh commands such as 'interface' group command, libvirt uses netcf library to directly mananger host network interface, it's unrelated with guest. Lucus and Chris, thanks for your suggests. Regards, Alex ----- Original Message ----- From: "Lucas Meneghel Rodrigues" <l...@redhat.com> To: "Chris Evich" <cev...@redhat.com> Cc: "Alex Jia" <a...@redhat.com>, g...@redhat.com, r...@redhat.com, autotest@test.kernel.org Sent: Wednesday, March 21, 2012 12:27:53 AM Subject: Re: [libvirt-autotest] New requirement On 03/20/2012 11:16 AM, Chris Evich wrote: > This would avoid adding another file to the tree, and keep the module > naming consistent. Thoughts? Ok, it's a sensible solution. I meant to work on the libvirt monitor for a while, but clearly we need to move out and have the code actually being used there. _______________________________________________ Autotest mailing list Autotest@test.kernel.org http://test.kernel.org/cgi-bin/mailman/listinfo/autotest
