headers
pcsol1996 | 1 Mar 08:07
Picon

Sun back to SPEC benchmarking with Linux again..

I remember when many of us were so happy when Sun started running it's
SPEC benchmarks on Solaris on it's x64 boxes..now not only is it back
to Linux, it isn't even using it's own compilers.  See below they used
Suse and Intel compilers.

Just another reason not to trust the Linux zealots (and those just
enamored of it) at Sun.

No excuse for this..

http://www.spec.org/cpu2006/results/res2007q4/cpu2006-20071015-02299.html

http://www.spec.org/cpu2006/results/res2007q4/cpu2006-20071015-02296.html

Please check the Links page before posting:
    http://groups.yahoo.com/group/solarisx86/links
Post message: solarisx86 <at> yahoogroups.com
UNSUBSCRIBE:  solarisx86-unsubscribe <at> yahoogroups.com
Permalink | Reply | 35 comments |
headers
Paul Floyd | 1 Mar 09:49
Picon
Favicon

Re: Sun back to SPEC benchmarking with Linux again..


On 1 Mar 2008, at 08:07, pcsol1996 wrote:

> I remember when many of us were so happy when Sun started running it's
> SPEC benchmarks on Solaris on it's x64 boxes..now not only is it back
> to Linux, it isn't even using it's own compilers.  See below they used
> Suse and Intel compilers.
>
> Just another reason not to trust the Linux zealots (and those just
> enamored of it) at Sun.
>
> No excuse for this..
>
> http://www.spec.org/cpu2006/results/res2007q4/cpu2006-20071015-02299.html
>
> http://www.spec.org/cpu2006/results/res2007q4/cpu2006-20071015-02296.html
>

Well, presumably the Intel compilers optimize better for Intel chips  
than do the Sun Studio compilers. And since there is no version of the  
Intel compilers for Solaris x86, it makes sense to publish the  
benchmarks with the fastest OS/compiler combination.

A+
Paul

Please check the Links page before posting:
    http://groups.yahoo.com/group/solarisx86/links
Post message: solarisx86 <at> yahoogroups.com
UNSUBSCRIBE:  solarisx86-unsubscribe <at> yahoogroups.com
(Continue reading)

Permalink | Reply | 0 comments |
headers
John Martin | 1 Mar 17:53
Picon

Re: Sun back to SPEC benchmarking with Linux again..

pcsol1996 wrote:
>
> I remember when many of us were so happy when Sun started running it's
> SPEC benchmarks on Solaris on it's x64 boxes..now not only is it back
> to Linux, it isn't even using it's own compilers. See below they used
> Suse and Intel compilers.
>
> Just another reason not to trust the Linux zealots (and those just
> enamored of it) at Sun.
>
>
> .
>
This Solaris zealot can only say this was a sensible choice at this 
point in time.
The Intel 10.1 compiler with build date 20070824 is producing very high
scores for libquantum, nearly 5x better scores than the earlier compilers.
The Studio compiler people are aware of the gap and are diligently trying
to solve it.

Please check the Links page before posting:
    http://groups.yahoo.com/group/solarisx86/links
Post message: solarisx86 <at> yahoogroups.com
UNSUBSCRIBE:  solarisx86-unsubscribe <at> yahoogroups.com
Permalink | Reply | 33 comments |
headers
Patch Nag | 2 Mar 03:35

New Solaris Patches available at riddleware.com 20080301

119255-52    SunOS 5.10_x86: Install and Patch Utilities Patch
119471-11    SunOS 5.10_x86: Sun Enterprise Network Array firmware and utilitie
119975-08    SunOS 5.10_x86: fp plug-in for cfgadm
120223-25    SunOS 5.10_x86: Emulex-Sun LightPulse Fibre Channel Adapter driver
120347-09    SunOS 5.10_x86: Common Fibre Channel HBA API and Host Bus Adapter
120544-11    SunOS 5.10_x86: Apache 2 Patch
125185-05    SunOS 5.10_x86: Sun Fibre Channel Device Drivers

Please check the Links page before posting:
    http://groups.yahoo.com/group/solarisx86/links
Post message: solarisx86 <at> yahoogroups.com
UNSUBSCRIBE:  solarisx86-unsubscribe <at> yahoogroups.com
Permalink | Reply | 0 comments |
headers
Al Hopper | 2 Mar 04:17

Re: Damn Small Solaris

On Fri, 29 Feb 2008, Mike Riley wrote:

> For those of you who heard of DSL (Damn Small Linux)... here's Damn
> Small Solaris :
>
> http://www.sunhelp.ru/archives/179-Damn_Small_Solaris_0.1.1_English_Page.html
>
> It helps if you can read Russian.

:)

After having been warned that the name might lead to trademark related 
issues, Alex has renamed the distribution to MilaX.  We're carrying a 
copy of the files at genunix.org.

Regards,

Al Hopper  Logical Approach Inc, Plano, TX.  al <at> logical-approach.com
            Voice: 972.379.2133 Fax: 972.379.2134  Timezone: US CDT
OpenSolaris Governing Board (OGB) Member - Apr 2005 to Mar 2007
http://www.opensolaris.org/os/community/ogb/ogb_2005-2007/

Please check the Links page before posting:
    http://groups.yahoo.com/group/solarisx86/links
Post message: solarisx86 <at> yahoogroups.com
UNSUBSCRIBE:  solarisx86-unsubscribe <at> yahoogroups.com
Permalink | Reply | 3 comments |
headers
pcsol1996 | 2 Mar 08:35
Picon

Re: Sun back to SPEC benchmarking with Linux again..

--- In solarisx86 <at> yahoogroups.com, John Martin <john.m.martin@...> wrote:
>
> pcsol1996 wrote:
> >
> > I remember when many of us were so happy when Sun started running it's
> > SPEC benchmarks on Solaris on it's x64 boxes..now not only is it back
> > to Linux, it isn't even using it's own compilers. See below they used
> > Suse and Intel compilers.
> >
> > Just another reason not to trust the Linux zealots (and those just
> > enamored of it) at Sun.
> >
> >
> > .
> >
> This Solaris zealot can only say this was a sensible choice at this 
> point in time.
> The Intel 10.1 compiler with build date 20070824 is producing very high
> scores for libquantum, nearly 5x better scores than the earlier
compilers.
> The Studio compiler people are aware of the gap and are diligently
trying
> to solve it.
>
Remember back several years ago with the old SPEC CPU, Sun had some
sort of "trick" with it's compilers (SPARC CPU) and one particular
benchmark that I can't recall now.  It became somewhat a joke since
everyone knew that the real performance of the processor was not
reflected.  I think anyone looking hard at SPEC CPU numbers would
ignore one extreme outlier of one benchmark.
(Continue reading)

Permalink | Reply | 32 comments |
headers
Gérard Henry | 2 Mar 12:50
Picon

Re: Damn Small Solaris

hello all,
just tried it with a usb key, and it 's nice on my dell laptop (already 
installed with SXDE 1/08).
It lacks the iwi driver, so i have to understand how to add it, hoping 
that it will fit on the usb key?
thanks to zfs, i can import my local disk, and have access to apps from 
blastwave. But i'm unable to launch firefox, because firefox needs 
dtksh. Am i wrong when i think that firefox can't run on opensolaris (it 
will be hacked i suppose?), because dtksh belongs to CDE, and CDE will 
never be shipped with opensolaris?

gerard

Please check the Links page before posting:
    http://groups.yahoo.com/group/solarisx86/links
Post message: solarisx86 <at> yahoogroups.com
UNSUBSCRIBE:  solarisx86-unsubscribe <at> yahoogroups.com
Permalink | Reply | 2 comments |
headers
Dick Hoogendijk | 2 Mar 14:24
Picon
Gravatar

dump to remote harddisk

My backup harddisk is on one of my local solaris boxes.
Now, I want to backup my server (S10u3) to that disk.

I guess (for safety reasons) I'd drop the server to single user
(downtime is no issue).

What would be the right thing to do next?
1) the NFS server/client model (dump to a nfs mounted disk)
or
2) dump through SSH to the remote (sxde4) box (which already has the
backup disk mounted).

I may be a stupid question for all of you've doen this often, but to me
it's kind of new. So, I welcome advise.

--

-- 
Dick Hoogendijk -- PGP/GnuPG key: 01D2433D
++ http://nagual.nl/ + SunOS sxde 01/08 ++

Please check the Links page before posting:
    http://groups.yahoo.com/group/solarisx86/links
Post message: solarisx86 <at> yahoogroups.com
UNSUBSCRIBE:  solarisx86-unsubscribe <at> yahoogroups.com
Permalink | Reply | 3 comments |
headers
Keith Bierman | 2 Mar 15:20
Picon
Gravatar

Re: Re: Sun back to SPEC benchmarking with Linux again..


On Mar 2, 2008, at 12:35 AM, pcsol1996 wrote:

>
> Remember back several years ago with the old SPEC CPU, Sun had some
> sort of "trick" with it's compilers (SPARC CPU) and one particular
>

The "trick" was a legitimate transformation.
>
> benchmark that I can't recall now. It became somewhat a joke since
> everyone knew that the real performance of the processor was not
>
The "real" performance of a processor is a combination of the  
hardware and software (especially on RISC systems, but even on CISC  
systems).
>
> reflected. I think anyone looking hard at SPEC CPU numbers would
> ignore one extreme outlier of one benchmark.
>
Sadly the vast majority of people don't look at the raw figures, they  
look exclusively at the summary number and make decisions based on it  
(silly, as an understanding of their actual workload and which  
component benchmarks actually correspond to their workload would  
enable much better decision making ... but it's almost never done.  
The vast majority of users have no idea what their workload(s) are,  
much less how to characterize them).
>
> Sun just needs to stick with Solaris and Studio 12.
>
(Continue reading)

Permalink | Reply | 31 comments |
headers
John D Groenveld | 2 Mar 15:56
Picon

Re: dump to remote harddisk

In message <20080302142409.00000d62 <at> westmark>, Dick Hoogendijk writes:
>What would be the right thing to do next?
>1) the NFS server/client model (dump to a nfs mounted disk)
>or
>2) dump through SSH to the remote (sxde4) box (which already has the
>backup disk mounted).

Both.

You can tunnel NFS over SSH, see Spencer Shepler's blog:
<URL:http://blogs.sun.com/roller/page/shepler?entry=tunneling_nfs_traffic_via_ssh>

You may want to play with ssh's compression and encryption options
depending on your available bandwidth and CPU.

Happy hacking,
John
groenveld <at> acm.org

Please check the Links page before posting:
    http://groups.yahoo.com/group/solarisx86/links
Post message: solarisx86 <at> yahoogroups.com
UNSUBSCRIBE:  solarisx86-unsubscribe <at> yahoogroups.com
Permalink | Reply | 2 comments |

Gmane