Re:Does XEN have a future?
on
Running Xen
·
· Score: 1
On CPU bound workloads, especially any CPU bound task that is "SMP aware"
The immense majority of real life task are not CPU bound anymore with current multi-core CPUS, but RAM and IO bound.
Re:Does XEN have a future?
on
Running Xen
·
· Score: 1
Anyone who pretends that Xen is supperior because of paravirtualization doesn't know what he is talking about. 1- VMware has paravirtualization also 2- paravirtualization only helps so much, and only with very recent linux hosts 3- paravirtualization is probably going to be useless (as in, not an improvement) once CPUs with virtualizable MMUs get common (shuch as lasts AMDs).
Suse's kernel version is only 2.4.21 inthe numbering, it's a very heavly pathched kernel.
With lot's of backports from 2.6 that are not on the vanilla 2.4.23.
(as a side effect it's was already pathched against the brk() bug when it was discovered )
Slashdot Mirror
Anyone who pretends that Xen is supperior because of paravirtualization doesn't know what he is talking about.
1- VMware has paravirtualization also
2- paravirtualization only helps so much, and only with very recent linux hosts
3- paravirtualization is probably going to be useless (as in, not an improvement) once CPUs with virtualizable MMUs get common (shuch as lasts AMDs).
Suse's kernel version is only 2.4.21 inthe numbering, it's a very heavly pathched kernel. With lot's of backports from 2.6 that are not on the vanilla 2.4.23. (as a side effect it's was already pathched against the brk() bug when it was discovered )