OpenBSD seems Much Faster Than FreeBSD -- Why? - BSD

This is a discussion on OpenBSD seems Much Faster Than FreeBSD -- Why? - BSD ; Hi! This is a strange problem. When I install FreeBSD during the install I tick (Developer) And install USERLAND. I then set optimization flags in /etc/make.conf :: cvsup and rebuild the world/userland with said flags. I then install xorg etc.. ...

+ Reply to Thread
Results 1 to 16 of 16

Thread: OpenBSD seems Much Faster Than FreeBSD -- Why?

  1. OpenBSD seems Much Faster Than FreeBSD -- Why?

    Hi!

    This is a strange problem. When I install FreeBSD during the install I
    tick (Developer) And install USERLAND. I then set optimization flags
    in /etc/make.conf :: cvsup and rebuild the world/userland with said
    flags. I then install xorg etc.. Everything is built from source, from
    xorg to firefox.
    Source based O/S should be faster than packaged rpms/debian etc.
    I install Gentoo, source mage, LFS from source and they don't run as
    fast as OpenBSD.

    I install OpenBSD and use their pre-built binaries/packages and it runs
    faster than Gentoo built from a stage1 tarball and FreeBSD built from
    source. Why Is that?


    I think I know why (I may be wrong) I ask a question in this group a
    few weeks ago. (My question was: How do you run PURE 64 IN FREEBSD)
    About a week later, some guy responds and tells me how. (I haven't
    tried I'm going to wait for FBSD-7 to hit RC-2. And give it a shot.


    My question is: Why is OpenBSD faster than a Gentoo system built from
    source and why is it faster than FreeBSD?

    TIA!









  2. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    In article ,
    Timmy writes:
    >
    > My question is: Why is OpenBSD faster than a Gentoo system built from
    > source and why is it faster than FreeBSD?
    >


    Faster at doing what?

    --
    Steve
    http://troutmask.apl.washington.edu/~kargl/

  3. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    On Wed, 30 Jan 2008 07:03:00 +0000 (UTC)
    kargl@troutmask.apl.washington.edu (Steven G. Kargl) wrote:

    > In article ,
    > Timmy writes:
    > >
    > > My question is: Why is OpenBSD faster than a Gentoo system built
    > > from source and why is it faster than FreeBSD?
    > >

    >
    > Faster at doing what?



    Faster at everything (on the side of a desktop system) Look, I don't
    have 'maple installed' and as for hardcore benchmarks running
    mathematical computations and this system finds a solution in an hour
    and this one in two hours etc. and bla bla.. Lookit, FreeBSD is my O/S.
    of choice, I like the fact that I have 18,000 programs in the ports
    tree, OpenBSD has about 4,000.

    OpenBSD seems faster than FBSD when it comes to simple things, for
    example starting firefox. Mail is also faster, using procmail,
    spamd/spamassassin, fetchmail, postfix and mutt.

    There is some cool things in OpenBSD that FBSD doesn't have.

    Steven G. Kargl do you want to 'chroot' or 'jail' your mail server?
    Allocating IP's and Jail/s is much easier in OpenBSD.

    OpenBSD is pretty cool when in comes to stuff like
    SSP/ProPolice etc.


    I'm not knocking freebsd in any shape or form.. I love FBSD that's my
    O/S of choice. What I don't get is this source based speed. LFS and
    Gentoo should be fast because you spent days compiling the system from
    source, same with FBSD. My point was, OpenBSD is the fastest O/S that
    I've used, and I didn't have to spend hours compiling. Don't know about
    you, but I'm starting to wonder about all of these source based O/S.
    Case in point. A friend of mine told me to check out Arch linux. So I
    did, it uses pacman/binaries and runs just as fast if not faster than
    gentoo. How I don't know..



  4. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    Timmy wrote:
    > On Wed, 30 Jan 2008 07:03:00 +0000 (UTC)
    > kargl@troutmask.apl.washington.edu (Steven G. Kargl) wrote:
    >
    > > In article ,
    > > Timmy writes:
    > > >
    > > > My question is: Why is OpenBSD faster than a Gentoo system built
    > > > from source and why is it faster than FreeBSD?
    > > >

    > >
    > > Faster at doing what?

    >
    >
    > Faster at everything (on the side of a desktop system)


    Look at this benchmark, it is old but says exactly the opposite of what
    you are saying:
    http://bulk.fefe.de/scalability/
    By the way, experience of most people is that OpenBSD may have
    qualities, byt speed is certainly not one of those. And in the years
    between the benchmark and now, FreeBSD and Linux have significantly
    improved their performance ...
    It may be that you experience is correlated to particular hardware you
    have which causes problems with FreeBSD.



    --

    Michel TALON


  5. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    On Wed, 30 Jan 2008 01:12:24 -0500, Timmy wrote:

    > I install OpenBSD and use their pre-built binaries/packages and it runs
    > faster than Gentoo built from a stage1 tarball and FreeBSD built from
    > source. Why Is that?


    I don't know, but if I were to hazard a guess from where I sit, it would
    be that you are comparing them on a relatively low-resource system, where
    OpenBSD's slightly smaller footprint gives it relatively more memory to
    bounce around in.

    On very low resource machines, NetBSD appears to run noticeably more
    quickly than OpenBSD for that reason.

  6. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    On Wed, 30 Jan 2008 08:38:53 +0000 (UTC)
    talon@lpthe.jussieu.fr (Michel Talon) wrote:

    > Timmy wrote:
    > > On Wed, 30 Jan 2008 07:03:00 +0000 (UTC)
    > > kargl@troutmask.apl.washington.edu (Steven G. Kargl) wrote:
    > >
    > > > In article ,
    > > > Timmy writes:
    > > > >
    > > > > My question is: Why is OpenBSD faster than a Gentoo system built
    > > > > from source and why is it faster than FreeBSD?
    > > > >
    > > >
    > > > Faster at doing what?

    > >
    > >
    > > Faster at everything (on the side of a desktop system)

    >
    > Look at this benchmark, it is old but says exactly the opposite of
    > what you are saying:
    > http://bulk.fefe.de/scalability/
    > By the way, experience of most people is that OpenBSD may have
    > qualities, byt speed is certainly not one of those. And in the years
    > between the benchmark and now, FreeBSD and Linux have significantly
    > improved their performance ...
    > It may be that you experience is correlated to particular hardware you
    > have which causes problems with FreeBSD.


    Noo, My hardware us up to pare.. X-2 amd64 2.4ghz with 4-gig of ram



    The guy wrote:

    "[Oct 20 2003] I was asked by a few OpenBSD people why I'm even
    comparing them here, since "everyone knows" they don't scale well and
    their goal is security and not scalability. The answer is: I didn't
    know that OpenBSD did not try to improve scalability as well. But even
    if I did know that, I would still have benchmarked them, if only to
    give people some numbers on how much less scalability they should
    expect. Knowing how much less scalability you get is also an important
    part of the decision process. And I also hope that security and
    scalability are not mutually exclusive and some volunteer veteran
    OpenBSD users will now be motivated to help port some scalability
    enhancements from FreeBSD and NetBSD to OpenBSD. My "fork crashes"
    problem was already reproduced and the OpenBSD people are working on
    it. That was fast! "
    ======================

    All I can tell ya Talon, is to install OBSD-4.2 and give it a try..
    It's very fast and responsive.

    I like FBSD better than OBSD - But OpenBSD is also a good operating
    system. I haven't tried NetBSD or Dragonfly.. As for OBSD all that
    built in cryptography/security is a +

  7. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    Hi Timmy,

    On Wed, 30 Jan 2008 03:01:43 -0500, Timmy wrote:

    > What I don't get is this source based speed.


    Not that I've got any particular reason to doubt your impressions, but
    you've offered exactly zero support for claims of one being faster than
    the other, or at what, or even any suggestion that the same tasks were
    being compared. So you must expect some scepticism.

    > LFS and Gentoo should be
    > fast because you spent days compiling the system from source, same with
    > FBSD.


    Well, that's not necessarily the case, any more. Once upon a time, in
    the 386/486/Pentium days [and even more so for the members of the RISC
    families] there were good reasons for doing target-specific
    optimization. This is *much less* the case with today's processors,
    because their multi-issue out-of-order nature, and sophisticated cache
    tweaks, pretty much do most of the fine-tune scheduling that you could
    want at run time. Sure, some numerical code can do with tweaking and
    ultra-optimization, but most operating-system code benefits from staying
    small (to get the most benefit from the i-cache) and not doing anything
    dumb. What I'm getting at, is that compiling from source is, these days,
    almost never because you want extra performance (there's none to be had),
    but usually because (a) it's easier in some way, (b) you want to have the
    source on hand so that you can debug easily, or (c) you're running odd-
    ball software on odd-ball hardware that no-one else cares about.

    > My point was, OpenBSD is the fastest O/S that I've used, and I
    > didn't have to spend hours compiling. Don't know about you, but I'm
    > starting to wonder about all of these source based O/S.


    You are right to wonder. But relax: you can get FreeBSD *and* the ports
    collection (most of it, anyway) in binary too. Be happy!

    [point: when you say you tweaked the compiler flags, did you do stuff
    like forcing in-lining, or loop unrolling? On a *lot* of OS-style code,
    those are actually pessimizations.]

    Cheers,

    --
    Andrew

  8. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    On 30 Jan 2008 09:20:07 GMT
    Andrew Reilly wrote:

    > Hi Timmy,
    >
    > On Wed, 30 Jan 2008 03:01:43 -0500, Timmy wrote:
    >
    > > What I don't get is this source based speed.

    >
    > Not that I've got any particular reason to doubt your impressions,
    > but you've offered exactly zero support for claims of one being
    > faster than the other, or at what, or even any suggestion that the
    > same tasks were being compared. So you must expect some scepticism.



    Like I said, I don't have any benchmarks, I have nothing but an
    opinion. OBSD seems faster than FBSD.

    > > LFS and Gentoo should be
    > > fast because you spent days compiling the system from source, same
    > > with FBSD.

    >
    > Well, that's not necessarily the case, any more. Once upon a time,
    > in the 386/486/Pentium days [and even more so for the members of the
    > RISC families] there were good reasons for doing target-specific
    > optimization. This is *much less* the case with today's processors,
    > because their multi-issue out-of-order nature, and sophisticated
    > cache tweaks, pretty much do most of the fine-tune scheduling that
    > you could want at run time. Sure, some numerical code can do with
    > tweaking and ultra-optimization, but most operating-system code
    > benefits from staying small (to get the most benefit from the
    > i-cache) and not doing anything dumb. What I'm getting at, is that
    > compiling from source is, these days, almost never because you want
    > extra performance (there's none to be had), but usually because (a)
    > it's easier in some way, (b) you want to have the source on hand so
    > that you can debug easily, or (c) you're running odd- ball software
    > on odd-ball hardware that no-one else cares about.


    Than what's the point of say; LFS Linux From Scratch?

    > > My point was, OpenBSD is the fastest O/S that I've used, and I
    > > didn't have to spend hours compiling. Don't know about you, but I'm
    > > starting to wonder about all of these source based O/S.

    >
    > You are right to wonder. But relax: you can get FreeBSD *and* the
    > ports collection (most of it, anyway) in binary too. Be happy!


    Yeah I know that. On my FBSD box I run portsnap fetch update and then
    portmaster -a


    > [point: when you say you tweaked the compiler flags, did you do stuff
    > like forcing in-lining, or loop unrolling? On a *lot* of OS-style
    > code, those are actually pessimizations.]


    I -fomit-frame-pointier & run funroll-loops on certain programs. For
    the most part -O2 -pipe.. I chated with this one women on Eggdrop who
    was a guru from MIT and she said these opto flags that I use cause the
    code to build a much larger code base .i.e. the more optimisations you
    use, the larger the code base, the larger the code base, the more time
    the pointers use/need to scan such code base and and load your program.
    Hmmm, I never thought about that way. She must be lying. From what I
    understand, you should throw every optimisation you can get by with. A
    larger optimised code base will run faster than a smaller unoptimised
    code base. If not we would all be build with CLFAGS -Oo -pipe. Or what
    ever the GCC options are to turn off optimisations ..

  9. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    Timmy wrote:
    >
    > Noo, My hardware us up to pare.. X-2 amd64 2.4ghz with 4-gig of ram
    >


    The problem is not the processor, it is the disk controller and similar
    stuff. I have seen several messages on FreeBSD mailing lists where some
    people have severe performance problems with FreeBSD which always seem
    correlated with very specific disk controllers or network cards. Of
    course these are the sort of problems the developers never see, and most
    users never see, so they are hard to solve. Since you have a biproc, SMP
    is a domain where OpenBSD should be particularly weak, according to
    *recent* benchmark i have seen but cannot find on the web now.


    --

    Michel TALON


  10. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    Begin
    On Wed, 30 Jan 2008 04:53:04 -0500, Timmy wrote:
    > On 30 Jan 2008 09:20:07 GMT
    > Andrew Reilly wrote:

    [snip!]
    >
    > Like I said, I don't have any benchmarks, I have nothing but an
    > opinion. OBSD seems faster than FBSD.


    You're offering nothing but your opinion. Great. What's worse, is that
    you're discounting other people's opinion based on that. You're not
    offering anything for discussion except for your convictions, so there's
    nothing to learn. You're even comparing apples and oranges and claiming
    it's alright because both are fruit.

    If this is not to be a troll, it's time you snapped out of it.


    [snip!]
    > Than what's the point of say; LFS Linux From Scratch?


    ``hack value''


    >> [point: when you say you tweaked the compiler flags, did you do stuff
    >> like forcing in-lining, or loop unrolling? On a *lot* of OS-style
    >> code, those are actually pessimizations.]


    Read this. Read it again. It clearly says ``OS-style code does not
    benefit from extensive optimizations, on the contrary''. There are
    reasons for this. And yet...


    > I -fomit-frame-pointier & run funroll-loops on certain programs. For
    > the most part -O2 -pipe..


    .... you use a bunch of those, and then:


    > I chated with this one women on Eggdrop who
    > was a guru from MIT and she said these opto flags that I use cause the
    > code to build a much larger code base .i.e. the more optimisations you
    > use, the larger the code base, the larger the code base, the more time
    > the pointers use/need to scan such code base and and load your program.


    You already have the answer on this. OS code has different needs from
    userland code. It runs shorter stretches at a time, surrounded by
    context switches, and thus benefits more from being smaller so it has a
    better chance of fitting in the instruction cache than from winning a
    few fractions by unrolling and avoiding branches.

    There, three people who said the same.


    > Hmmm, I never thought about that way. She must be lying. From what I
    > understand, you should throw every optimisation you can get by with. A
    > larger optimised code base will run faster than a smaller unoptimised
    > code base.


    .... you rant about how someone from mit must clearly be wrong because
    of what Andrew just said, too.

    Timmy, you're way out of your depth, you're not making sense, and
    you're loudly ignoring people who try to explain it to you.

    Would you mind?


    > If not we would all be build with CLFAGS -Oo -pipe. Or what
    > ever the GCC options are to turn off optimisations ..


    Optimisations are no silver bullets, Timmy. Exploiting circumstances
    doesn't work if the circumstances you're relying on change.


    --
    j p d (at) d s b (dot) t u d e l f t (dot) n l .
    This message was originally posted on Usenet in plain text.
    Any other representation, additions, or changes do not have my
    consent and may be a violation of international copyright law.

  11. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    Timmy wrote:
    > was a guru from MIT and she said these opto flags that I use cause the
    > code to build a much larger code base .i.e. the more optimisations you
    > use, the larger the code base, the larger the code base, the more time
    > the pointers use/need to scan such code base and and load your program.
    > Hmmm, I never thought about that way. She must be lying.


    She is right. The single most important factor is the size of the L2 and
    L1 caches. If you code can run in cache it runs *much* faster than if
    you need to access memory. If, after optimisation (which generally
    increases codes size, due to loop unrolling etc.) the code still fits in
    cache, then it may run a little faster thanks to the optimisation. But
    if it doesn't fit any more in cache, then it will run slower. If it is
    so big that only a small part fits in cache, then most of the time is
    spent accessing memory, and only experience shows if optimisation is of
    any benefit.

    > From what I
    > understand, you should throw every optimisation you can get by with. A
    > larger optimised code base will run faster than a smaller unoptimised
    > code base. If not we would all be build with CLFAGS -Oo -pipe. Or what
    > ever the GCC options are to turn off optimisations ..


    Yes, and -O has been the recommended flag for a very long time by
    FreeBSD developers.

    --

    Michel TALON


  12. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    On Wed, 30 Jan 2008 10:25:49 +0000, Michel Talon wrote:
    > Timmy wrote:
    >> From what I
    >> understand, you should throw every optimisation you can get by with.


    No. "Optimization" is really a misnomer for most of the things that a
    compiler can or does do. "Heuristic" is probably a better term. There
    really aren't many situations where loop-unrolling is a net win these
    days, for example. The termination condition and branch in conventional
    code cost essentially zero, because of the parallel execution and branch
    prediction mechanisms. But the code space increase that causes a single
    extra cache miss causes the processor to sit and twiddle it's thumbs for
    hundreds of cycles: possibly 1000 or more missed instruction slots. That
    could represent the entire execution path of a system call.

    >> A
    >> larger optimised code base will run faster than a smaller unoptimised
    >> code base. If not we would all be build with CLFAGS -Oo -pipe. Or what
    >> ever the GCC options are to turn off optimisations ..


    Not quite: GCC generates particularly stupid code at -O0, (or at least
    used to.) Jumps to jumps and that sort of thing. -O (aka -O1) is
    usually a fair recipe for reasonable code generation. Look at the gcc
    manual: it includes quite a lot of well-rounded heuristics.

    > Yes, and -O has been the recommended flag for a very long time by
    > FreeBSD developers.


    True, but now that they've moved to GCC-4, (at least on amd64) the
    recommendation is -O2. Presumably that's been found to work the best. -
    O2 doesn't do most of the heroic/pessimising optimizations, but it does
    do as much as it can of the good register allocation and the usual common
    subexpression stuff. It would be mildly interesting to compare binary
    sizes and execution speed on some sort of benchmark, at -O1 and -O2, but
    benchmarking well is hard (particularly operating systems), and I can't
    be bothered. Anyone who does: post the results, please?

    Cheers,

    --
    Andrew

  13. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    Michel Talon wrote:

    > Timmy wrote:
    >>
    >> Noo, My hardware us up to pare.. X-2 amd64 2.4ghz with 4-gig of ram
    >>

    >
    > The problem is not the processor, it is the disk controller and similar
    > stuff. I have seen several messages on FreeBSD mailing lists where some
    > people have severe performance problems with FreeBSD which always seem
    > correlated with very specific disk controllers or network cards. Of
    > course these are the sort of problems the developers never see, and most
    > users never see, so they are hard to solve. Since you have a biproc, SMP
    > is a domain where OpenBSD should be particularly weak, according to
    > *recent* benchmark i have seen but cannot find on the web now.
    >
    >


    There are just too many details missing. I am not trying to discount
    subjective "impressions", as when I see a difference large enough to not
    require benchmark hair-splitting I know it when I see it.

    We simply do not know if, per chance, his disk controller (or NIC, or USB)
    may be using a driver that is still GIANT locking, or perhaps there is a
    hidden "interrupt storm" he hasn't noticed, etc...

    I'm not even clear on whether we are talking about 6.3 or 7.0. Most of the
    performance related scalability improvements I've been following are mostly
    related to work on the ULE scheduler in conjunction with the libthr
    threading library. This is really only relevant to 7.0, and then only if
    the system has been configured to use them. ULE is not the default
    scheduler yet, even though I believe this is slated to possibly change for
    7.1-Release.

    We don't know if he is using a PAE kernel, or 64-bit with 32-bit
    compatibility libraries. Either will not perform as well as native.

    So far, I have two machines at home running 7.0 RC1. Both are UP, built with
    nooption KSE and utilizing ULE/libthr. Since they are freshly installed I
    also have no COMPAT options at all built into the kernel either. All the
    userland ports were linked against only 7.0 version libraries. I may
    attempt this at work with a dual Xeon in the next week, or so, if I get
    time.

    No benchmarks yet - but "subjectively" speaking they both seem to be doing
    well, in that I don't get the "impression" either are slower than the
    6.2-Release they were running previously. The only item to surface so far
    has been related to mod_php under Apache segfaulting [worker-mpm and we
    know some of PHP isn't thread safe], and that has been narrowed down to the
    eaccelerator opcode cache. My guess here is a quirk with sysvshm. The
    interesting thing is it doesn't appear on the second box which is running
    lighttpd with PHP as a FastCGI.

    So much here we just don't know...

    -Jason


  14. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    Jason Bourne wrote:

    [snip]
    >
    > We don't know if he is using a PAE kernel, or 64-bit with 32-bit
    > compatibility libraries. Either will not perform as well as native.
    >


    PS:

    I also meant to include debug kernel, symbols, Invariants, Witness, etc.
    He also may have built without stripping symbols as well. All known
    performace robbers.

    [snip]
    > -Jason



  15. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    Timmy wrote:
    >Steven G. Kargl wrote:
    >>> My question is: Why is OpenBSD faster than a Gentoo system built
    >>> from source and why is it faster than FreeBSD?

    >>
    >> Faster at doing what?

    >
    > Faster at everything (on the side of a desktop system) Look, I don't
    > have 'maple installed' and as for hardcore benchmarks running
    > mathematical computations and this system finds a solution in an hour
    > and this one in two hours etc. and bla bla.. Lookit, FreeBSD is my
    > O/S. of choice, I like the fact that I have 18,000 programs in the
    > ports tree, OpenBSD has about 4,000.


    User perception is something that is important, but it also very hard to
    quantify and measure. That's precisely why "hard numbers" are used when
    comparing two systems.

    > OpenBSD seems faster than FBSD when it comes to simple things, for
    > example starting firefox. Mail is also faster, using procmail,
    > spamd/spamassassin, fetchmail, postfix and mutt.


    This may be true, but just because you say it is true doesn't make it
    so. Please take the time to *measure*.

    > I'm not knocking freebsd in any shape or form.. I love FBSD that's my
    > O/S of choice. What I don't get is this source based speed. LFS and
    > Gentoo should be fast because you spent days compiling the system from
    > source, same with FBSD.


    That's where the problem lies. The assumption that ``I compiled this
    from source, so it should be faster'' is at best a bit naive, and at
    worst completely stupid.

    There are _many_ options which you can use to optimize binaries for a
    particular system's hardware, software workload, and projected use in
    the future. The sheer number of these options is enough to confuse the
    hell out of someone with only cursory knowledge of a few `fancy'
    compiler options. This great number of options, and their combinations
    too, are a constant source of misunderstanding and false expectations.

    Many people believe that compiling with gcc -O3 ``will make their codes
    run faster''. That may be true. Just not always, and not for all sorts
    of programs or libraries, or not on all sorts of hardware.

    Since the assumption ``a source-based system is faster'' is often false,
    the rest of the conclusions based on this assumption (about FreeBSD
    vs. Gentoo stage1 vs. OpenBSD vs. Arch Linux vs. whatever is `in' this
    month) are also of questionable value.

    Before you `optimize' a particular system for a particular workload, it
    is worth sitting down and describing in detail the parts of the system:
    both the hardware *and* the software it runs. *Then* it is time to
    start gathering hard numbers. There's a reason why people insist on
    seeing these things. They are repeatable, measurable, and quantifiable
    in a way that user perception never really is.

    - Giorgos


  16. Re: OpenBSD seems Much Faster Than FreeBSD -- Why?

    In article <1aWnj.4392$jj6.221@newsfe06.lga>,
    Timmy writes:
    > On Wed, 30 Jan 2008 07:03:00 +0000 (UTC)
    > kargl@troutmask.apl.washington.edu (Steven G. Kargl) wrote:
    >
    >> In article ,
    >> Timmy writes:
    >> >
    >> > My question is: Why is OpenBSD faster than a Gentoo system built
    >> > from source and why is it faster than FreeBSD?
    >> >

    >>
    >> Faster at doing what?

    >
    > Faster at everything (on the side of a desktop system) Look, I don't
    > have 'maple installed' and as for hardcore benchmarks running
    > mathematical computations and this system finds a solution in an hour
    > and this one in two hours etc. and bla bla.. Lookit, FreeBSD is my O/S.
    > of choice, I like the fact that I have 18,000 programs in the ports
    > tree, OpenBSD has about 4,000.


    You've given us zero content to try to help identify how you
    may have misconfigured FreeBSD or problems with the hardware
    (ie, interrupt storms). Heck, you haven't even told us which
    version of FreeBSD your using nor given an adequate description
    of the hardware.

    Are you using softupdates on your filesystems? Is /etc/malloc.conf
    set?

    What are the compilers on the different systems (ie gcc --version)?
    This is very important in that gcc-3.4.x produces in general better
    code than gcc-4.2.x.

    Try reading the tuning manpage.

    > Steven G. Kargl do you want to 'chroot' or 'jail' your mail server?


    No.

    >
    > I'm not knocking freebsd in any shape or form.
    >


    I understand. Unfortunately, you're also not giving us any
    information that can help identify possible problems.

    --
    Steve
    http://troutmask.apl.washington.edu/~kargl/

+ Reply to Thread