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

RE: [Xen-devel] highly available xen

>> >. migrate a domain in the case of an orderly shutdown
>> Not sure I understand this one.

>Think he meant to auto-migrate domains off a physical server when that
server is deliberately taken >down for some reason.

> >. restart a domain on another physical server in the case of a 
> >complete xen crash

> >. add some smarts to know how many domains a physical server can 
> >support

> >. prioritise critical domains over non-critical domains (eg a primary

> >mail server might take priority over a secondary dns server)

>These are the sort of things I think could usefully go into a
higher-level "cluster controller".  We >identified a few requirements
for a cluster controller some time ago but nobody here has had the
>spare cycles required to implement it.

I implemented this exact scenario using drbd, heartbeat and tripwire.
Drbd handles the replication of domU metadata to the secondary physical
server. Heartbeat checks that the primary is 'alive' and handles
resource takeovers in a failover event. Tripwire performs periodic tests
on the dom0 and will force migration if a serious error occurs.

The point is, this is all doable using already written and more
importantly mature tools that have done their time in the trenches. If
we are going to bother implementing this sort of thing at the Xen level
then we had better make it better than whats already there, or else we
are wasting our time.


SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
Xen-devel mailing list



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