>
>
> 1. How to specify a VMID to resume in?
> Can you specify which vmid to resume in?
> For example, what I want to do is like this?
> # xm save 1 debian.chk
> # xm restore debian.chk 1 <- here you usually resume this in dom2,
> not 1 as it used to be.
> I want resume debian.chk in dom1!.
>
>
> 2. About Xen live migration where both PC have same disk image.
> Assume PC_A and PC_B both have very same /XEN/Debian.img, but
> inode num, type of FS, etc,,, are different.
> I ask you that if I do live-migration fromm PC_A to PC_B,
> can I resume the image without using iSCSI or NBD?
> I mean, when the VM stays on PC_A, it mounts /XEN/Debian.img on
PC_A''s
> HDD,
> and after it migrates to PC_B, it mounts /XEn/Debian.img on PC_B HDD.
> Is this possible?
AFAIU, this is not possible. The file handles opened by the OS on PC_A
correspond to inodes on the file system on PC_A. When the VM is migrated to
PC_B, the inodes refered in the file handles are no longer found (or are
wrong inodes). So the VM will hang or give some unpredictable behavior. It
is necessary that the storage should be visible at the device node level on
PC_A and PC_B, hence ISCSI or AOE or cluster FS will be needed.
3. How to fork "after resumed function" inside
VMs?> Is this possible to do suspending from inside of DomU?
> It is like this:
> [Dom3]# /bin/suspend.sh
> ---------------------
> suspend.sh:
> #!/bin/sh
> /usr/sbin/xm save /mnt/NFS/dom0/XEN/$DOM_NAME/suspend.img
> [here suspends]
> /etc/init.d/dhcpc restart
> [After resumed, you get dhcp re-run automatically.]
> ---------------------
>
>
>
> _______________________________________________
> Xen-users mailing list
> Xen-users@lists.xensource.com
> http://lists.xensource.com/xen-users
>
--
Jayesh
------------------------------------------------------------------------
Everything you can imagine is real
_______________________________________________
Xen-users mailing list
Xen-users@lists.xensource.com
http://lists.xensource.com/xen-users