[lug] virtualization first boot question
karl horlen
horlenkarl at yahoo.com
Sat Sep 18 11:43:39 MDT 2010
> On 09/14/2010 05:49 PM, karl horlen wrote:
> >1) when i reboot my server with the xen kernel, [...] i
> don't have to
> >screw with [...]
>
> My experience with Xen has been that using "xen" and "don't
> have to screw
> with" within close proximity is almost never correct.
although i was looking for the simplest way to do this, as i stated in the other post, ease of use isn't the end and be all. if a given implementation is a little more difficult but works well once it's configured, imho it's worth learning and committing to. if it's a pita and a poor implementation, that's a different story though. i'm not sure you're saying that xen is the latter, just that you're going to have to dig into some manuals.
> I'd almost consider VMWare server for your situation as
> being low-stress,
> but that's not perfect either. Apparently there's
> some screwing around you
> have to do to get the console browser working in current
> Linuxes or
> something, and the performance suspending and resuming
> virtuals can be
> terrible. It probably will run directly on top of
> your existing host with
> the least screwning around and breaking of the current
> setup though.
the thing i worry about with vmware server is that i believe it's outside of centos' yum and proprietary which i fear will relate to some kind of dependency purgatory or compatibility issues at some point. also, i wonder about features. i think different versions or feature implementations of vmware server are commercial, depending on what you need. i'd like to go down a path where i don't get roped into buying something down the road.
is it generally easy to migrate guests from different implementations these days, just in case you decide to change paths in the future? you'd obviously have reconfigure the networking, but i'm asking more about migrating the underlying disk images / partitions to a new scheme.
sean what virtual server technology do you use mostly?
thanks
More information about the LUG
mailing list