On Tue, Feb 28, 2012 at 2:50 PM, Ken D'Ambrosio <k...@jots.org> wrote:
> There *are* distributed filesystems -- btrfs has ceph, which has come a > long, > long way. Lustre and Gluster also come to mind. Caveat: I've not used > these, > I worked with the good folks at Gluster at my last job, fall of 2009. They were great, but it just did not work well for my application (storing virtual machine boot images) but I expect it will work very well for this application. It was tricky to setup at the time, but they were just about to launch a major release that was supposed to make everything trivial, so maybe that's the way to go. > On Tue, 28 Feb 2012 14:44:54 -0500 Kenny Lussier wrote > > I thought about drbd, but that is active/backup only. > I'm not sure what you are getting at here, but I use DRBD as part of an HA storage stack (DRBD>LVM>iSCSI>Heartbeat) for cloud storage in the host machines. I believe this would work nicely for your purposes as well. Add in heartbeat to take care of failover and you could be good to go. I picked pieces from the very well written Building a redundant iSCSI and NFS cluster with Debian<http://www.markround.com/archives/44-Building-a-redundant-iSCSI-and-NFS-cluster-with-Debian-Part-1.html>to setup HA storage on Ubuntu Lucid hosts and I'm near the finish line of a reboot of the project at my current employer using CentOS 6 hosts. Don't be scared off by all the source builds in that guide. It is all done with a package manager now. Just a matter of figuring out which packages/repos to install. <http://www.markround.com/archives/44-Building-a-redundant-iSCSI-and-NFS-cluster-with-Debian-Part-1.html>
_______________________________________________ gnhlug-discuss mailing list gnhlug-discuss@mail.gnhlug.org http://mail.gnhlug.org/mailman/listinfo/gnhlug-discuss/
