On Nov 27, 2007 2:15 PM, carlopmart <[EMAIL PROTECTED]> wrote: > Hi all, > > I have setup a rhel5.1 xen server on production environment running 5 > windows > guests installed over iscsi disks using lvm to store OS windows and data. But > now, which is it the best form for doing hot bakcups of whole disks and reduce > minimal downtime like, for example, vmware esx does with snaphosts ??
Right now I'm using a script that: - uses 'xm save' to pause the domain and write out a memory file - uses an lvm snapshot for the volume containing the disk image - restores the saved memory file - backs up the memory file and disk snapshot - removes the snapshot In my tests, I was able to restore the memory file and disk image on another server and start the domain; for some odd reason, moving the VM to another machine makes the time off by some 5 days on the VM, but I can just immediately reboot the VM to get the correct time. The total downtime for the VM is about 25s for 512MB ram, and 52s for 1024MB. I really haven't checked to see how long a full shutdown/snapshot/startup would take, but my guess is this is faster. Also, this allows some long-lived connections to remain open. YMMV, but I'm pretty happy with this solution. Regards, David -- David L. Parsley Linux Systems Engineer, James Madison University "If I have seen further, it is by standing on ye shoulders of giants" - Isaac Newton _______________________________________________ rhelv5-list mailing list [email protected] https://www.redhat.com/mailman/listinfo/rhelv5-list
