Description of problem: When I start virt-manager and chose File|New machine and chose an nfs installation source, the installation lasts forever. How reproducible: always Steps to Reproduce: 1. start virt-manager 2. chose File|New machine -> -> System1 -> paravirtualized -> enter an NFS installation source -> Simple File -> Shared physical device -> -> Finish Actual results: Installation lasts forever Expected results: Installation does not last forever Additional info: Every second, I get on the console: libvir: error : library call virConnectNumOfDefinedNetworks failed, possibly not supported libvir: error : library call virConnectNumOfNetworks failed, possibly not supported
I think the message every second and the installation time may be unrelated. I did install over an NFS mounted partition in the past, it was slower than local disk install due to network access but worked, I would probably need to retry. For the message could you make sure you have the libvirt daemon running, it is started by '/etc/init.d/libvirtd start' and should have spun off a dnsmasq instance too ? Also for NFS install, please make sure you checked the box asking for initialization of the disk space, it does a zeroing allocation of the file used for the volume before starting the new domain, which may look a bit long initially, but avoid really costly reallocation routines while the filesystem grows during installation. It precisely avoided such case of very long installation times we hit if the volume wasn't initialized ahead of time. If that's not the problem, could you specify at what point the installation starts to stall ? Daniel
Thanks to your hints, the error message every second no longer appears. Thanks. Now I see I am asked for a password in the console. As soon as I enter whatever as password, the installation stops with the console error message: [root@ls3099 ~]# virt-manager Password: mount error 5 = Input/output error Refer to the mount.cifs(8) manual page (e.g.man mount.cifs) umount: /var/lib/xen/virtinstmnt.ypwgWQ: not mounted and I get the graphical error msg: Unable to complete install 'exceptions.RuntimeError Invalid install location Traceback (most recent call last): File "/usr/share/virt-manager/virtManager/create.py", line 681, in do_install dom = guest.start_install(False, meter = meter) File "/usr/lib/python2.4/site-packages/virtinst/Guest.py", line 647, in start_install tmpfiles = self._prepare_install_location(meter) File "/usr/lib/python2.4/site-packages/virtinst/ParaVirtGuest.py", line 65, in _prepare_install_location (kernelfn,initrdfn,args) = DistroManager.acquireKernel(self.location, meter, scratchdir=self.scratchdir, type=self.type) File "/usr/lib/python2.4/site-packages/virtinst/DistroManager.py", line 566, in acquireKernel raise RuntimeError, "Invalid install location" RuntimeError: Invalid install location '
I am now using a normal file, no nfs location, and no network. Now it works.
I have now performed the installation procedure exactly as the first time, but only replaced the nfs://vol/what/ever/path/x86_64/RHEL5.1-Server-20071017.0-x86_64-DVD.iso by /where/nfs/is/mounted/x86_64/RHEL5.1-Server-20071017.0-x86_64-DVD.iso and it works.
Ahhh, okay, I misunderstood the problem. Virt-manager (and in general all the stack) will manipulate NFS files transparently using their file paths as they are seen when mounted. Trying to mount filesystem on the fly raises the problem of the authentication and associated user interface, it's better to have them pre-mounted (or automatically available via automount). I assume the problem is solved at this point, if it's not the case please reopen explaining what is still blocking you, Daniel