eyck wrote:
> You can get away with running without those:
>
>> CAP_SYS_MODULE
>> CAP_MKNOD
>>
> hmm, actually without those too:
>
>> CAP_NET_ADMIN
>> CAP_NET_RAW
>>
>
> (You just need to load vmmon and vmnet on the host, there needs to be vmnet
> configuration on the host that matches the one that guest uses, and then
> all you need is some kind of X-server on the guest to run vmware
> player/server)
>
That means running them on the host, and my idea was to have as
absolutely as little as possible on the host.
My original comments *were* the short version. A slightly longer
version, for example, is that you can get away without things like
CAP_MKNOD, but only if you manually create the nodes that the script
tries to do. I started with that, but got tired of re-running it for
every node it failed for. I'm guessing I could revoke the capability
now, under the assumption that, once installed, VMware won't want to
alter the nodes.
I *think* I need the CAP_NET_* caps because of the way VMware wants to
do networking. I haven't exhausted the possibilities here, so you might
be right.
> Both xen and esx use hypervisor, I guess they wouldn't like running one on
> top of the other...
>
The Xen and QEMU comments were a joke. My wife doesn't get my jokes
either. But in seriousness, I'm using the (free (beer)) VMware Server,
not ESX / GSX, and this VMware Server doesn't need a hypervisor.
On that, I wonder if the vserver patches and Xen patches can co-exist.
I shudder to think of the patching. But it might be nice to have them
both in one - the flexibility of Xen (different kernels etc) and the
efficiency of vservers.
Tony
_______________________________________________
Vserver mailing list
Vserver@list.linux-vserver.org
http://list.linux-vserver.org/mailman/listinfo/vserver
Received on Fri Apr 21 11:31:29 2006