[rescue] Cheap, fast, PC server

Robert Darlington rdarlington at gmail.com
Sun Jan 27 21:42:04 CST 2008


Can I chat with you off line about this sometime?  This year I'm going
to implement a fail over cluster for hosting all our existing servers
in a virtualized environment (be it with VMWare, Windows Virtual
Server, or a combination of the two.  iSCSI storage on the back and
and diskless booting of the cluster is the goal (internal disks for
swap only).


On Jan 27, 2008 8:20 PM, Bill Bradford <mrbill at mrbill.net> wrote:
> On Sun, Jan 27, 2008 at 09:06:13AM -0800, Devin Ganger wrote:
> > If you're doing everything all in the same OS, yes. However, if you've got a
> > variety and want to maximize hardware use while minimizing power and cooling
> > load, VMs are the way to go. VMs also have the advantage of being a lot
> > easier to move onto a different server in the event of hardware issues --
> > simply have one backup process on the host to appropriately suspend and
> > protect your VM files, and another backup process on the guest to capture
> > the data that's changed (treating the two much like a full and a
> > differential, respectively).
> At the $BIGCORP I work at, most of the Exchange infrastructure and AD
> server farms are on virtual machines.  If a VM node goes down, the VMs just
> get automatically migrated to another system; this prevents a single
> hardware failure from taking out a large portion of infrastructure.
> We're also using a 3-node ESX cluster in my department to get rid of tons
> of "legacy" boxes; stuff like Dell Precision 220s that should have been
> tossed 5-6 years ago but yet still run "critical" services (at least in
> the opinions of some people).
> Bill
> --
> Bill Bradford
> Houston, Texas
> _______________________________________________
> rescue list - http://www.sunhelp.org/mailman/listinfo/rescue

More information about the rescue mailing list