At Thu, 06 Oct 2011 23:16:09 -0700, Mark Pace wrote: > Thanks for the info on the collie commands. > > I think the key advantage to having a "canonical" backup store would be that you didn't have to stress the collie cluster with backup requests because the data would already be on that one unit. With the data multicasting around, it seems a shame to run up the cluster for a redundant request if we could create the node that stores it all by default. But the backup node limits the total storage size of Sheepdog, doesn't it? I think we cannot create a large cluster with the architecture. > > 2. Images/snapshots to files: a mechanism to take a snapshot or a > > regular image out of the sheepdog cluster to turn it into a file > > for backup, to be sent to a customer, stored on a hard drive, etc. > > This may already exist in qemu-img, but I don't know how easy it is > > to take a snap and turn it into a full image. This is more of a one > > off thing than the first point, but is necessary none-the-less. I > > see that qcow2 images can be created from snapshots using a > > relatively new (or a least the site I read claimed it was relatively > > new) feature of the qemu-img convert functionality -- does sheepdog > > also support this? No, Sheepdog doesn't support it now. I'll create a patch to support it after QEMU block maintainer comes back from vacation. But you can do the similar thing with collie as Frank said. Thanks, Kazutaka |