[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[Xen-users] [XCP] Concrete recommendation XCP 1.0/1.1 and experience w/ fault tolerant storage - e.g. DRBD

Dear List,

first of all kudos for XCP community effort! As a long time Xen user I
think this is one of the most interesting development in years.
As we like to migrate our existing Xen infrastructure to XCP, we are in
need of HOWTOs and best practices  - I do not know if this slightly
off-topic, but anyways I try:

What we would like to achieve is the following based on two XCP
- LVM based aproach based on a volume group of distributed phyisical
volumes via DRBD
- Allocate block devices as logical volumes
- Share the DomU configuration between both XCPs

Hence when a VM gets created on one XCP installation the VM is always
availble at the other XCP installation (both configuration and data).
When one server is down the other machine takes over the active role (if
not already) and the VM can be booted on that host.
If there is a sheduled maintenance of one machine XCP should also be
able to do a life migration?

While we are not determined on using DRBD it seems stable mature and
reasonable for our task at hand. Our current mirroring is based on
rsyncing lvm snapshots, so we do not have much experience with DRBD.

I found some pointers on installing DRBD on XCP (albeit 0.5):

There are also other recent threads mentioning DRBD as a solution (e.g.
as a replacement for VastSky):

So we would be very interested in:
- Pointers to concrete steps for using DRBD on XCP (1.0/1.1).
- Hints and pitfalls when going this route
- (Better) alternative approaches (that are not too cutting edge)
- Best practices for configuration of fault tolerant XCP installations
(e.g. syncing configuration among XCP instances, automatic failover?)

Thanks in advance,

Xen-users mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.