Hi, My setup is arguably smaller than yours, so YMMV:
Key Point: I have found that using infrastructure provided natively by Solaris/ZFS are the best choices. I have been using CIFS... it's unpredictable when some random windows machines would stop seeing them. XP/Server 2003/Vista - Too many things go wrong. So here is what I do: 1. I use xVM VirtualBox for Windows 2. Snapshot/Clones are managed by ZFS... Just put the vmdk on its own FS and let zfs handle all sorts of shiny stuff. In your case same thing can be done with ZFS Volumes, if you choose to go with iSCSI. 3. There is a reason whole industry relies on NFS (learned the hard way). In short - it works! I have just installed SFU on all windows clients and couldn't be happier. No matter what happens to server or network (unplugging the wrong cable or router rebooting) , the clients *always* behave predictably. When the server/network goes up, it all again starts working. I do *all* shares through NFS now - windows, Linux and Solaris. Easy to setup *and* predictable! In short, for minimum fuss, stick with letting the bottom most layer which makes sense manage stuff... and in any case, don't let the virtualization products manage the storage in _any_ way. ZFS does it best and let it do it... and stick with NFS for sharing. SFU NFS client/server are small and they work really well, and they come bundled with Windows Server 200x. For 8 drives, go to raidz2. Two drives going bad is easy in an 8 drive setup. -- This message posted from opensolaris.org _______________________________________________ zfs-discuss mailing list zfs-discuss@opensolaris.org http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
