List Info

Thread: Re: xen 3.1 problem (Re: xen 3.1.0 is there)




Re: xen 3.1 problem (Re: xen 3.1.0 is there)
user name
2007-06-28 18:00:57
   On Jun 28, 18:54, Manuel Bouyer wrote:
   > Subject: Re: xen 3.1 problem (Re: xen 3.1.0 is
there)
   > On Thu, Jun 28, 2007 at 10:44:20PM +0900, Kazushi
Marukawa wrote:
   > > It's a part of idutils package, devel/idutils. 
This mkid
   > > parses all .c, .h, .s files and creates a tag
file named
   > > 'ID'.  It is not a benchmark program but it
acceses files
   > > and it spends about 10 secs once files are
cached on 512M
   > > memory allocated for DomU.  I'm not sure it is
suitable to
   > > evaluate the system or not for our case,
though.  It was
   > > just handy.  
   > 
   > OK, so here's what I get. All this with a dom0
kernel from today + my
   > interrupt patches:
   > 4.99.21                   8.54 real         7.63
user         0.89 sys
   > 4.99.21 + my patches:     8.52 real         7.39
user         1.06 sys
   > current 06220000Z         8.43 real         7.47
user         0.94 sys
   > 4.0_BETA2                 8.33 real         7.49
user         0.81 sys

Is it possible to run the same program on domU too?

I pointed that domU with recent kernel runs slower than
dom0.  The dom0 runs as the same all the time.  

Regards,
-- Kazushi
In seeking the unattainable, simplicity only gets in the
way.
		-- Epigrams in Programming, ACM SIGPLAN Sept. 1982

Re: xen 3.1 problem (Re: xen 3.1.0 is there)
country flaguser name
France
2007-06-29 03:23:24
On Fri, Jun 29, 2007 at 08:00:57AM +0900, Kazushi Marukawa
wrote:
>    On Jun 28, 18:54, Manuel Bouyer wrote:
>    > Subject: Re: xen 3.1 problem (Re: xen 3.1.0 is
there)
>    > On Thu, Jun 28, 2007 at 10:44:20PM +0900,
Kazushi Marukawa wrote:
>    > > It's a part of idutils package,
devel/idutils.  This mkid
>    > > parses all .c, .h, .s files and creates a
tag file named
>    > > 'ID'.  It is not a benchmark program but
it acceses files
>    > > and it spends about 10 secs once files are
cached on 512M
>    > > memory allocated for DomU.  I'm not sure
it is suitable to
>    > > evaluate the system or not for our case,
though.  It was
>    > > just handy.  
>    > 
>    > OK, so here's what I get. All this with a dom0
kernel from today + my
>    > interrupt patches:
>    > 4.99.21                   8.54 real        
7.63 user         0.89 sys
>    > 4.99.21 + my patches:     8.52 real        
7.39 user         1.06 sys
>    > current 06220000Z         8.43 real        
7.47 user         0.94 sys
>    > 4.0_BETA2                 8.33 real        
7.49 user         0.81 sys
> 
> Is it possible to run the same program on domU too?

This was all on domU. My comment on dom0 was to point out
that all the
tests were run with the same dom0 kernel (otherwise there
would be 16
results instead of 4 

-- 
Manuel Bouyer, LIP6, Universite Paris VI.          
Manuel.Bouyerlip6.fr
     NetBSD: 26 ans d'experience feront toujours la
difference
--

[1-2]

about | contact  Other archives ( Real Estate discussion Medical topics )