Am Mittwoch, den 20.02.2013, 15:16 +0100 schrieb Paolo Bonzini: > Il 20/02/2013 14:46, Kevin Wolf ha scritto: > >> > Thanks for the fast reply. Can you please elaborate on that (or post a > >> > link to an explanation): What exactly does the backend have to provide > >> > that for example cache=writeback can be used? Is there a way to test for > >> > it? > > The problem is about cache coherency. Local files work just fine with > > cache=writeback, but if you migrate to a different host, you get the > > problem that the destination kernel may cache some parts of the image > > while the source is still writing to it. When you do the actual switch > > to the destination host and the backend doesn't take care of > > invalidating that cache, you might (partially) read outdated data. > > Also, note that some backends (iSCSI and NBD for example) will never use > a cache on the host, so they can be used with the default > cache=writeback mode. > > However, this is simply because writeback is mapped to none, and > writethrough to directsync. If you used iSCSI or NBD through the > kernel, you would need cache=none.
We are using GlusterFS. Currently with a FUSE-based mount with qcow2s on it but switching to qemu-GlusterFS-devices later. But thanks for the clarifications. Tiziano -- stepping stone GmbH Neufeldstrasse 9 CH-3012 Bern Telefon: +41 31 332 53 63 www.stepping-stone.ch tiziano.muel...@stepping-stone.ch
