Re: [Xen-devel] [xen-unstable test] 128240: regressions - FAIL

[Dario Faggioli <dfaggioli@xxxxxxxx>]

On Tue, 2018-10-02 at 10:36 +0100, Jan Beulich wrote:
> > > > On 02.10.18 at 11:24, <dfaggioli@xxxxxxxx> wrote:
> > 
> No. See Jürgen's response. The default scheduler (irrespective of
> whether it was chosen via command line option) should not matter.
> Any means to force a non-default scheduler (and it indeed looks
> like CPU pools are the only way) should imo retain that specific
> scheduler. 
>
Right, but the scheduler is a Xen/host (or cpupool) property, not a VM
one. Do we handle like that the other similar ones? Like, do we fail
migration if one host use the default setting wrt autoballooning and
the other doesn't? Or wrt XPTI and the other mitigations?

I mean, I think I see your point, and I'm still making up my mind about
this, but I'm starting to think that this is the user responsibility to
know what he's doing, and all we should do is, in this case, if we
notice the mismatch between the schedulers, to print a warning and
avoid setting the scheduler params...

Regards,
Dario
-- 
<<This happens because I choose it to happen!>> (Raistlin Majere)
-----------------------------------------------------------------
Dario Faggioli, Ph.D, http://about.me/dario.faggioli
Software Engineer @ SUSE https://www.suse.com/

Attachment: signature.asc
Description: This is a digitally signed message part

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel