Mailing List Archive

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Peter A. H. Peterson wrote:
> Hi Everyone,
>
> My name is Peter Peterson and I represent a group of a grad students
> at UCLA. We're in a computer systems performance analysis course and we
> were hoping to do a general performance comparison of gentoo vs. a
> popular binary i386-compatible distribution (probably ubuntu) in some
> "real-world" server tests to try and meaningfully calculate the
> performance gains that local compilation provides. (For example,
> apache2 requests processed per second on the same hardware.)
>
> I've subscribed to this list because we want the gentoo community to
> be involved in helping us design the tests so that we can hopefully
> all feel good about what and how we are testing the systems.
>
> We have no particular outcome in mind; our group represents a wide
> range of computer users, from Mac, Linux, and Windows enthusiasts, and
> we have all used a wide variety of Linux distributions. We have simply
> noticed that much of the discussion of gentoo's performance advantage
> is anecdotal and we're genuinely hoping to provide some meaningful
> experimental data for discussion. Also, if anyone knows of any
> available benchmark data or papers on this subject, we'd love to hear
> about them. There was apparently a paper on slashdot a couple of years
> ago, but the host it was on appears to now be squatted. For that
> matter, if this is a well understood or closed issue (for example, if
> the statistics that people quote are actually from good experimental
> data) please let us know.
>
> Is anyone here interested in discussing this project? We are
> specifically interested in discussing methodology, testing suits,
> CFLAGS and other options. Our desire is not to "trick out" gentoo or
> ubuntu, but rather quantify the performance benefit that gentoo has
> over binary distributions with "normal" compile flags (whatever normal
> is).
>
> Thanks for your time,
>
> Peter Peterson (et al)
>

The one thing I know is that the Ruby interpreter gains quite a bit
(about 20 - 30 percent) from being compiled with the appropriate
"-march=" flag in GCC, and that most of the other compilation options
have much less payoff. I don't have the link handy, but it certainly
makes sense.

Today's chips are optimized to run lots of streams of 386 code, so a
"server" test might not show substantial differences. If you've got
enough threads to keep the processor(s) busy, you aren't going to see
much of an advantage from compiling to "native" code. Where you will see
a pronounced advantage is on single-threaded applications, like
scientific workstations run.

Quite frankly, I'd be surprised if a Gentoo server was significantly
faster than a CentOS 5 (RHEL 5 clone) server on a high-intensity server
workload. But I have tried a lot of distros for scientific workstations,
and Gentoo does seem to have an advantage there.

What's the title of the course? Is this a compiler course?
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGNB1LaZUb+jwczfoRAprsAKC08NiyZSNnzOV2vAYIEgShQnFRigCgzddR
6htoqVXzJE2lSb6R5+0ftZA=
=GFdy
-----END PGP SIGNATURE-----
--
gentoo-performance@gentoo.org mailing list

> Quite frankly, I'd be surprised if a Gentoo server was significantly
> faster than a CentOS 5 (RHEL 5 clone) server on a high-intensity
> server workload. But I have tried a lot of distros for scientific
> workstations, and Gentoo does seem to have an advantage there.

I would have to agree. The performance gains in Gentoo tend to (In the
Real World(tm) ) come from not having a bunch of crap instaled an
running that hogs memory rather than from using a particular set of
optimizing flags. Further, real-word servers ten to be built with some
performance margin in order to handle load spikes, so a few percent
here and there wouldn't really be measurable, assuming you compare to a
system that isn't bloated.

Further,this list has averaged at about one post per month since around
2003 when I joined, so don't expect to gett too many replies from here.

--DA

I have to agree with Daniel here. This list is very low traffic. If you want
to reach a broad part of the gentoo community, I think the Gentoo forums
could be the way to go. ( http://forums.gentoo.org ). As for the performance
tests, I don't think either that the server world would give you a lot of
differences in terms of performance whether you're using Gentoo or a popular
server-oriented binary distro.

Just my 2 cents...

On 4/29/07, Daniel Armyr <daniel.armyr@home.se> wrote:
>
> > Quite frankly, I'd be surprised if a Gentoo server was significantly
> > faster than a CentOS 5 (RHEL 5 clone) server on a high-intensity
> > server workload. But I have tried a lot of distros for scientific
> > workstations, and Gentoo does seem to have an advantage there.
>
> I would have to agree. The performance gains in Gentoo tend to (In the
> Real World(tm) ) come from not having a bunch of crap instaled an
> running that hogs memory rather than from using a particular set of
> optimizing flags. Further, real-word servers ten to be built with some
> performance margin in order to handle load spikes, so a few percent
> here and there wouldn't really be measurable, assuming you compare to a
> system that isn't bloated.
>
> Further,this list has averaged at about one post per month since around
> 2003 when I joined, so don't expect to gett too many replies from here.
>
> --DA
>
>

Hi..

Your experience Daniel it is so good, I the company what I work have the
SunFire V440 with CentOS and It is so bad, so bad, actually one of the most
important problamens what we have is :

One day the company page can not be accessed from anywahere, including the
same company, and the apache (httpd), network services seems to be fine, we
probing so many things, the last what we try was restart al services related
with networking, and guess what, the problem was resolved. Looking in the
web for causes of the problem, I read in a web CentOS have so problems,
CentOS sometimes said "the service es OK" but that is not true.

Few days later we (thecnicians) have a meeting to talk about the problem,
and we decide migrate to gentoo, when the deparment coordinator talk with
the boss, the boss said NO! gentoo NO!, because CentOS it is more
productive, and gentoo is so much dificult to configure and our cordinator
said "no boss it is not true" and the boss said "this is my last word".

I think what the boss decision is no good because Gentoo it is better, much
better in many ways.

What I want with all that?... Have any people in the list some experiences
with this kind of migrate? somebody can tell me about your experience?

Thanks in advance and regards..

Note : Excuse me for my english, I am learning... Thanks and Excuses... Have
a nice day

On 4/29/07, Guillaume Ceccarelli <lastrainson@gmail.com> wrote:
>
> I have to agree with Daniel here. This list is very low traffic. If you
> want to reach a broad part of the gentoo community, I think the Gentoo
> forums could be the way to go. ( http://forums.gentoo.org ). As for the
> performance tests, I don't think either that the server world would give you
> a lot of differences in terms of performance whether you're using Gentoo or
> a popular server-oriented binary distro.
>
> Just my 2 cents...
>
> On 4/29/07, Daniel Armyr <daniel.armyr@home.se> wrote:
> >
> > > Quite frankly, I'd be surprised if a Gentoo server was significantly
> > > faster than a CentOS 5 (RHEL 5 clone) server on a high-intensity
> > > server workload. But I have tried a lot of distros for scientific
> > > workstations, and Gentoo does seem to have an advantage there.
> >
> > I would have to agree. The performance gains in Gentoo tend to (In the
> > Real World(tm) ) come from not having a bunch of crap instaled an
> > running that hogs memory rather than from using a particular set of
> > optimizing flags. Further, real-word servers ten to be built with some
> > performance margin in order to handle load spikes, so a few percent
> > here and there wouldn't really be measurable, assuming you compare to a
> > system that isn't bloated.
> >
> > Further,this list has averaged at about one post per month since around
> > 2003 when I joined, so don't expect to gett too many replies from here.
> >
> > --DA
> >
> >
>


--
Francisco Rivas

It seems I'm still on this list. Although it's low traffic, some interesting
stuff comes across. I don't have anything useful to add, but I'd be interested
in following the progress of your group. If the information is going to be made
public or if you are starting a forum thread, drop a line here if you don't
mind. Good luck.
--
gentoo-performance@gentoo.org mailing list

Yes, i think having formal results would be so useful to a lot of
people, including a lot of us that are tring to convince our bosses to
migrate to gentoo =)

I reforce the asking of lnxg33k, I am interessed in following the
progress of your group too.

By the way, I am working on a migration of a NT PDC to linux. I am
doing all the tests in gentoo box for now, but probabily my boss will
ask to set up the production server in a debian.
So, if I can be of some help, let me now.

ps.: excuse for my english, still learning.
Fabiano.

On 4/29/07, lnxg33k <lnxg33k@gmail.com> wrote:
> It seems I'm still on this list. Although it's low traffic, some interesting
> stuff comes across. I don't have anything useful to add, but I'd be interested
> in following the progress of your group. If the information is going to be made
> public or if you are starting a forum thread, drop a line here if you don't
> mind. Good luck.
> --
> gentoo-performance@gentoo.org mailing list
>
>


--
Fabiano.
--
gentoo-performance@gentoo.org mailing list

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Daniel Armyr wrote:
>> Quite frankly, I'd be surprised if a Gentoo server was significantly
>> faster than a CentOS 5 (RHEL 5 clone) server on a high-intensity
>> server workload. But I have tried a lot of distros for scientific
>> workstations, and Gentoo does seem to have an advantage there.
>
> I would have to agree. The performance gains in Gentoo tend to (In the
> Real World(tm) ) come from not having a bunch of crap instaled an
> running that hogs memory rather than from using a particular set of
> optimizing flags. Further, real-word servers ten to be built with some
> performance margin in order to handle load spikes, so a few percent
> here and there wouldn't really be measurable, assuming you compare to a
> system that isn't bloated.
>
> Further,this list has averaged at about one post per month since around
> 2003 when I joined, so don't expect to gett too many replies from here.
>
> --DA

Except, of course, from a professional Linux performance engineer who
runs Gentoo scientific workstations at home. :)
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGNNNOaZUb+jwczfoRAh71AKC4XN2z1vAqpk+ireB0IuNFZeYFuACgwjre
YiQqQ3ddLvbOhUwR/GcuATg=
=Bwoa
-----END PGP SIGNATURE-----
--
gentoo-performance@gentoo.org mailing list

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Francisco Rivas wrote:
> Hi..
>
> Your experience Daniel it is so good, I the company what I work have the
> SunFire V440 with CentOS and It is so bad, so bad, actually one of the most
> important problamens what we have is :
>
> One day the company page can not be accessed from anywahere, including the
> same company, and the apache (httpd), network services seems to be
> fine, we
> probing so many things, the last what we try was restart al services
> related
> with networking, and guess what, the problem was resolved. Looking in the
> web for causes of the problem, I read in a web CentOS have so problems,
> CentOS sometimes said "the service es OK" but that is not true.
>
> Few days later we (thecnicians) have a meeting to talk about the problem,
> and we decide migrate to gentoo, when the deparment coordinator talk with
> the boss, the boss said NO! gentoo NO!, because CentOS it is more
> productive, and gentoo is so much dificult to configure and our cordinator
> said "no boss it is not true" and the boss said "this is my last word".
>
> I think what the boss decision is no good because Gentoo it is better, much
> better in many ways.
>
> What I want with all that?... Have any people in the list some experiences
> with this kind of migrate? somebody can tell me about your experience?
>
> Thanks in advance and regards..
>
> Note : Excuse me for my english, I am learning... Thanks and Excuses...
> Have
> a nice day
Since this is a Gentoo list and not a CentOS list, email me off-list for
further discussion. But what you describe sounds like the random
glitches any "24x7" web server will have, regardless of which distro you
use. It's why Red Hat Enterprise Linux is a profitable *product* and Red
Hat Certified Engineers exist, and why until other distros have that
kind of muscle, they won't be seriously considered in such migrations.

CentOS is kind of a special case. Their aim is to produce an RHEL
"clone" -- rebuild is the correct term. For the most part, they are only
a day or so behind RHEL if that. So many if not most of the issues you
ran into would have also happened on RHEL.
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2.0.3 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFGNNU6aZUb+jwczfoRAnVjAKC9eO7WYuAdTcAdtAWuaIiJv9vrxwCgjWvM
rZB0ZX8z0nhYue3kSgpqio4=
=6rfd
-----END PGP SIGNATURE-----
--
gentoo-performance@gentoo.org mailing list

Hi there,

On 4/29/07, Peter A. H. Peterson <pedro@tastytronic.net> wrote:
> Hi Everyone,
>
> My name is Peter Peterson and I represent a group of a grad students
> at UCLA. We're in a computer systems performance analysis course and we
> were hoping to do a general performance comparison of gentoo vs. a
> popular binary i386-compatible distribution (probably ubuntu) in some
> "real-world" server tests to try and meaningfully calculate the
> performance gains that local compilation provides. (For example,
> apache2 requests processed per second on the same hardware.)
>
> I've subscribed to this list because we want the gentoo community to
> be involved in helping us design the tests so that we can hopefully
> all feel good about what and how we are testing the systems.
>
> We have no particular outcome in mind; our group represents a wide
> range of computer users, from Mac, Linux, and Windows enthusiasts, and
> we have all used a wide variety of Linux distributions. We have simply
> noticed that much of the discussion of gentoo's performance advantage
> is anecdotal and we're genuinely hoping to provide some meaningful
> experimental data for discussion. Also, if anyone knows of any
> available benchmark data or papers on this subject, we'd love to hear
> about them. There was apparently a paper on slashdot a couple of years
> ago, but the host it was on appears to now be squatted. For that
> matter, if this is a well understood or closed issue (for example, if
> the statistics that people quote are actually from good experimental
> data) please let us know.
>
> Is anyone here interested in discussing this project? We are
> specifically interested in discussing methodology, testing suits,
> CFLAGS and other options. Our desire is not to "trick out" gentoo or
> ubuntu, but rather quantify the performance benefit that gentoo has
> over binary distributions with "normal" compile flags (whatever normal
> is).
>

A good CFLAGS would be something not very agressive, something like:
-march=<cpuType> -O3 or -O2 and at most -fomit-frame-pointer.
(Scientific workloads can speedup considerably with: -ffast-math)

Having experienced and done some benchmarks with gentoo and other
distros on servers and on scientific workstations.
What I found is that sometimes gentoo lacks critical performance
patches in glibc that are applied to mainstream distros (redhat,
suse..etc) that provide boosts in memcpy, memset, etc..(I remmember a
discussion about that some years ago).
What I also found out is that the compiler flags only affect
workloads that are very compute intensive. not something that depends
almost completely on FSB load or IO load.. like most server
workloads... -O3 doesn't do much to a working set full of
unpredictable branches (like server workloads usually are) and low IPC
rate.

I really do believe performance boost from gentoo to be practically
negligible. The difference will only be apreciable in very few corner
cases. Most distros also optimize critical aplications such has:
openssl, mplayer.. reducing the possible corner cases.

Anyways, doing a "academic" benchmark would be a good idea.

Something like:
micro-benchmarks:
- stream (mem bandwith benchmark)
- ??

macro-benchmarks:
- apache2 + gzip + php(make it cpu intensive, not IO intensive)
- xmlmark ?
- kernelbench
- pybench ?
- openssl bench


about methodology:
- same system, same bios version, same disks.
- All OSes must be installed in the same disk partitions.
- the will be trouble about the kernel config:
- for mainstream distros you should use the kernel that is provided.
- for gentoo, gentoo-sources configured by someone which is
experienced, and informed about configuration impacts on performance
(ideally a kernel hacker?).

- should use the stable versions in gentoo portage?
- or should use the same application versions used on mainstream distro?



--
Miguel Sousa Filipe
--
gentoo-performance@gentoo.org mailing list

I'd agree with Miguel. In general, I can't tell the difference in most
applications. I did some tests on OpenOffice a few years ago and found the 8
hour compile time wasn't worth it in terms of performance. Where Gentoo
shines is in giving you the increased ability to shrink your installed size.
I have nothing on this machine that I don't specifically want there and I
work overly hard to keep my compile flags exactly right.

That said, I'd be interested in seeing your results. It would be nice to
know that all of this work is beneficial for more than just keeping me
happy.

Mike

On 4/30/07, Miguel Sousa Filipe <miguel.filipe@gmail.com> wrote:
>
> Hi there,
>
> On 4/29/07, Peter A. H. Peterson <pedro@tastytronic.net> wrote:
> > Hi Everyone,
> >
> > My name is Peter Peterson and I represent a group of a grad students
> > at UCLA. We're in a computer systems performance analysis course and we
> > were hoping to do a general performance comparison of gentoo vs. a
> > popular binary i386-compatible distribution (probably ubuntu) in some
> > "real-world" server tests to try and meaningfully calculate the
> > performance gains that local compilation provides. (For example,
> > apache2 requests processed per second on the same hardware.)
> >
> > I've subscribed to this list because we want the gentoo community to
> > be involved in helping us design the tests so that we can hopefully
> > all feel good about what and how we are testing the systems.
> >
> > We have no particular outcome in mind; our group represents a wide
> > range of computer users, from Mac, Linux, and Windows enthusiasts, and
> > we have all used a wide variety of Linux distributions. We have simply
> > noticed that much of the discussion of gentoo's performance advantage
> > is anecdotal and we're genuinely hoping to provide some meaningful
> > experimental data for discussion. Also, if anyone knows of any
> > available benchmark data or papers on this subject, we'd love to hear
> > about them. There was apparently a paper on slashdot a couple of years
> > ago, but the host it was on appears to now be squatted. For that
> > matter, if this is a well understood or closed issue (for example, if
> > the statistics that people quote are actually from good experimental
> > data) please let us know.
> >
> > Is anyone here interested in discussing this project? We are
> > specifically interested in discussing methodology, testing suits,
> > CFLAGS and other options. Our desire is not to "trick out" gentoo or
> > ubuntu, but rather quantify the performance benefit that gentoo has
> > over binary distributions with "normal" compile flags (whatever normal
> > is).
> >
>
> A good CFLAGS would be something not very agressive, something like:
> -march=<cpuType> -O3 or -O2 and at most -fomit-frame-pointer.
> (Scientific workloads can speedup considerably with: -ffast-math)
>
> Having experienced and done some benchmarks with gentoo and other
> distros on servers and on scientific workstations.
> What I found is that sometimes gentoo lacks critical performance
> patches in glibc that are applied to mainstream distros (redhat,
> suse..etc) that provide boosts in memcpy, memset, etc..(I remmember a
> discussion about that some years ago).
> What I also found out is that the compiler flags only affect
> workloads that are very compute intensive. not something that depends
> almost completely on FSB load or IO load.. like most server
> workloads... -O3 doesn't do much to a working set full of
> unpredictable branches (like server workloads usually are) and low IPC
> rate.
>
> I really do believe performance boost from gentoo to be practically
> negligible. The difference will only be apreciable in very few corner
> cases. Most distros also optimize critical aplications such has:
> openssl, mplayer.. reducing the possible corner cases.
>
> Anyways, doing a "academic" benchmark would be a good idea.
>
> Something like:
> micro-benchmarks:
> - stream (mem bandwith benchmark)
> - ??
>
> macro-benchmarks:
> - apache2 + gzip + php(make it cpu intensive, not IO intensive)
> - xmlmark ?
> - kernelbench
> - pybench ?
> - openssl bench
>
>
> about methodology:
> - same system, same bios version, same disks.
> - All OSes must be installed in the same disk partitions.
> - the will be trouble about the kernel config:
> - for mainstream distros you should use the kernel that is provided.
> - for gentoo, gentoo-sources configured by someone which is
> experienced, and informed about configuration impacts on performance
> (ideally a kernel hacker?).
>
> - should use the stable versions in gentoo portage?
> - or should use the same application versions used on mainstream distro?
>
>
>
> --
> Miguel Sousa Filipe
> --
> gentoo-performance@gentoo.org mailing list
>
>

Miguel Sousa Filipe wrote:
>
> A good CFLAGS would be something not very agressive, something like:
> -march=<cpuType> -O3 or -O2 and at most -fomit-frame-pointer.
> (Scientific workloads can speedup considerably with: -ffast-math)

You must be careful on cpuflags. -ffast-math should never be used
globally. Applications that benefit from it specify it themselves. The
most realistic is -O2.

The remark about processors being optimized for i386 is a bit off beat.
Actually what -march does is enable additional instructions to be used
that are faster (and available) on those processors. Especially mmx{,2},
sse{,2,3} are very beneficial to certain kinds of applications, and can
give big speedups. Considering that, however, many applications that are
sensitive to this such as mplayer can do run-time cpu detection and use
different code (often written in assembler) depending on the cpu at runtime.

In general it is not wise to use -O3. Most distributions compile with
-O2. -O3 therefore is less well tested and may expose bugs in both the
compiler and (more importantly) the applications.

Besides the instructions that are available to the compiler (influenced
by -march) there is then the issue of instruction scheduling. To put it
simply, all modern processors have the possibility to process multiple
instructions in parallel when they do not conflict. The precise
architectures of this differ and the ability to parallelize depends
quite much on the scheduling of the instructions. This is influenced by
the -mtune flag (set implicitly by -march). This is an optimization that
is most likely also present in a binary distribution. Gentoo can however
optimize to the precise processor used.

Paul

--
gentoo-performance@gentoo.org mailing list

Quoting lnxg33k:
> It seems I'm still on this list. Although it's low traffic, some
> interesting stuff comes across. I don't have anything useful to add, but
> I'd be interested in following the progress of your group. If the
> information is going to be made public or if you are starting a forum
> thread, drop a line here if you don't mind. Good luck.

Hey all,

Just to let you know, we are progressing on our project, and I did
start a post on the gentoo forums (but I will keep checking here).
I've been too busy to post, but I really appreciate your thoughts.

We will definitely make our results public.

Here's the forum:

http://forums.gentoo.org/viewtopic-t-556107.html

Thanks again!

Peter

--
Peter A. H. Peterson, technician and musician.
---=[ http://tastytronic.net/~pedro/ ]=---
--
gentoo-performance@gentoo.org mailing list

A good topic, I want be a part of this, Peter or Paul can you tell me how to
be part of your projects?...

Thank you very much in advance for your answer...

On 5/10/07, Peter A. H. Peterson <pedro@tastytronic.net> wrote:
>
> Quoting lnxg33k:
> > It seems I'm still on this list. Although it's low traffic, some
> > interesting stuff comes across. I don't have anything useful to add, but
> > I'd be interested in following the progress of your group. If the
> > information is going to be made public or if you are starting a forum
> > thread, drop a line here if you don't mind. Good luck.
>
> Hey all,
>
> Just to let you know, we are progressing on our project, and I did
> start a post on the gentoo forums (but I will keep checking here).
> I've been too busy to post, but I really appreciate your thoughts.
>
> We will definitely make our results public.
>
> Here's the forum:
>
> http://forums.gentoo.org/viewtopic-t-556107.html
>
> Thanks again!
>
> Peter
>
> --
> Peter A. H. Peterson, technician and musician.
> ---=[ http://tastytronic.net/~pedro/ ]=---
> --
> gentoo-performance@gentoo.org mailing list
>
>


--
Francisco Rivas
Linux User (New) : #448324
Linux Machine (New) : 355187