W dniu 2013-08-02 17:28, Trond Endrestøl pisze: > On Fri, 2 Aug 2013 17:04+0200, ?ukasz W?sikowski wrote:
>> Thank you for you reply. I did what you've suggested, now I'm getting: >> >> ZFS: i/o error - all block copies unavailable >> ZFS: can't read object set for dataset u >> ZFS: can't open root filesystem >> gptzfsboot: failed to mount default pool klawisz >> >> Those KVM VMs are very stubborn with this "I won't boot" attitude ;) > > Could there be a bug in the underlying storage system? It might be a bug in Proxmox or KVM. > What filesystem does the VM's virtual harddrives reside on? VM uses qcow container which is placed on ext4 on top of software raid-10. > Could the physical storage media be corrupt? No. I created identical VM on another physical Proxmox node and got the same boot error. > For some reason ZFS expects additional (leaf) vdevs to be present. > I've seen similar error messages when an insufficient number of drives > were not made visible by the boot firmware, i.e. BIOS. The most odd thing is that this setup is working for a while and then stops. > If this VM is important, you should make a snapshot, or make a copy, > if you have enough free space. I can prepare another VM with that booting problem for tests so I'm not afraid of data loss. > And it might be best to seek advice from someone more versed with the > internals of ZFS than I am. :-/ Thank you for trying :) -- best regards, Lukasz Wasikowski _______________________________________________ [email protected] mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-stable To unsubscribe, send any mail to "[email protected]"
