Re: High availability on two boxes

2008/5/5, Gijs <info boer-software-en-webservices nl>:

You might want to take a look at http://www.drbd.org/.
It basically does what you suggest. It creates a software raid 1 partition across a local and networked partition.
Perfect for real-time syncronisation between disks across a network, while not having to resort to expensive stuff like SANs and the like.

Regards,

Gijs

Uno Engborg wrote:

Hi,
I'm experimenting with a HA setup on two physical boxes. The idea is to create virtual Xen machines that automagically do live migration from one physical box to the other in case of e.g. a hardware failure on one of the physial boxes.

My idea is to create a software raid of networked block devices and real disks. I.e:

Box A:
Software raid consisting of:
Physical disk in box A
Nework block device exported from box B

Box B:
Software raid consisting of:
Physical disk in box B
Network block device exported from box A

On top of this I plan t use GFS2 and a cluster that handles the migration of the virtual machines.

Any comments on this, would this work, or would it result in deadlock? What about quorum disk? I suppose
I need one? Most setup of these kind of things usually have their storage on a separate
SAN, but as this is mostly for experimentation and testing purposes, I would hope that I could do without that for now.

What would be the best way to export the network block devices? I'm thinking iSCSI or GNDB?

If this doesn't look like a good idea, are there any other way to do this.

Regards
Uno Engborg

--
fedora-list mailing list
fedora-list redhat com
To unsubscribe: https://www.redhat.com/mailman/listinfo/fedora-list