Slashdot Mirror
Open Source OS Benchmarking Competition
BenchmarkingFreak writes "OSnews is running a story about a new benchmarking competition: OSU Open Source Lab wanted to take the concept of benchmarking a little bit further with the Beaver Challenge 2004. In this competition they will be allowing a community of experts in each OS to tweak their configurations to ensure maximum performance. And they are running it all on wicked machines, just imagine... well you know."
that is going to investigate bush. Somethings wrong here.
I use Gentoo; how does this affect me??
FP. Out.
* Debian GNU/Linux
* Fedora Linux
* FreeBSD
* Gentoo Linux
* NetBSD
* OpenBSD
* Red Hat Linux
* Slackware Linux
* SuSE GNU/Linux
Where's Mandrake?
this was already done with Win2k3 Datacenter edition vs. Linux.
Win2k3 R0X0red Linux's World!!!11!!!
Gartner ROX!
First a story about screws, now a story about beavers. Apparently the /. crew had a slow weekend.
It will be decided once and for all that Gentoo offers no discernable advantage over more conventional distros. At which point, Gentoo-zealots will only have portage to pull in converts.
And portage is being... ported to Slackware
Discuss here!
Anonymous due to obviousness of replay (not like karma whoring is even an issue anymore)
where are the white women at?
Funny, I can't seem to find the source for Xandros out on their website. Isn't that a condition of using the Linux kernal?
of course, debian will lose.
what did you expect?
Will they be benchmarking database performance, GCC compiling speed, I took at look at the methodology page and it wasn't particularly specific.
LINDOWZ RULEZ ALL
I just heard some sad news on talk radio - Horror/Sci Fi writer Stephen King was found dead in his Maine home this morning. There weren't any more details. I'm sure everyone in the Slashdot community will miss him - even if you didn't enjoy his work, there's no denying his contributions to popular culture. Truly an American icon.
This is a really neat idea, and it's a long time coming; however, I wouldn't expect overly divergent results among Linux distributions. Afterall, they're all going to use the latest 2.4 and 2.6 kernel and comparable glibc versions(with maybe the exception of Debian), so the only speed difference should be in the compiler flags used to build the packages. I'm not trying to negate the coolness of this competition because it should give a good measure of performance between the BSD distros VS Linux distros, but don't be surprised when the Linux distros all show comparable results. As a footnote, I do expect Gentoo to come in the lead of the Linux distros having tried them all and found it the fastest in empircal testing...
A musician without the RIAA, is like a fish without a bicycle.
I hear they all take vacation after the superbowl!
How do I keep track of people who are fingering
Mandrake is more of a "user friendly" distro then a performance optimized distribution. Someone might add Mandrake to the list, but it's not as tightly configured as say Gentoo and I can't imagine what purpose it would be to add it to the competition except for just representation.
If they really wanted to make it a good competition, they would award beaver to the winner. That would get them fired up.
They're benchmarking the BSD's vs. Linux. If Linux wins, the BSD snobs will have a field day babbling and flaming about how unfair the contest was. Beware the BSD assholes!
Some girls don't like it quick.
Given OSNews' recent penchant for poorly-done benchmarks (e.g. 1, 2), I'm glad to see them run an article about someone else's (hopefully well-done) testing. By having expert teams who know what they're doing tweak the configurations, this should be a much more representative result. Hopefully OSNews will learn some methodology from these guys...
"You can never have too many elephants on your team."
With all this talk of beavers and tweaking i might just have to go out and catch me some hot case porn.. Newegg Here i come!
Sig (appended to the end of comments you post, 120 chars)
The only way to get real world benchmark results would be to distribute the benchmarks setiathome style and then do some funky statitical analysis. But you'd also need a way to verify the actual hardware used for each test.
Otherwise, we still end up with data that needs to be heavily interpreted to get any take on what will happen in production on any given hardware.
Oh yeah, not to mention that for real world production, performance is also dependant on maintainability, uptime and a variety of other factors.
Still, this will result in a bunch of pretty graphs and nice rants and raves and dick waving for years to come, even though the data will only be good for one given point in time, for one given hardware platform, for one given configuration.
There's no way in hell that a geek is going to win any "Beaver Challenge."
I just heard the sad news on CBC radio. Comedy actor/writer Alan Thicke was found dead in his home this morning. Even if you never liked his work, you can appreciate what he did for 80's television. Truly a Canadian icon. :(
He will be missed
Show me That Smile (The Growing Pains Theme Song):
Show me that smile again.
Ooh show me that smile.
Don't waste another minute on your crying.
We're nowhere near the end.
We're nowhere near.
The best is ready to begin.
As long as we got each other
We got the world
Sitting right in our hands.
Baby rain or shine;
All the time.
We got each other
Sharing the laughter and love.
"Ask not what your country can do for you." --John F. Kennedy
This is daft!
It's like me choosing a Renault over a Jaguar just because they woop their ass in Formula 1. Tweaking the OS on a souped up server is not the same as using an out of the box distro on bog standard hardware.
What's the bet the 'tweaking' involves recompiling every component with hardware optimised CFLAGS etc.?
They will use 2 GB of RAM.
At 896 MB and below, Linux can use some
extra optimizations. During kernel config,
you choose the "1 GB" option.
Above 8 GB, and especially around 32 GB,
you switch to the 4g/4g model. On very large
systems, FreeBSD won't even boot because
some data structures become too large for
the low memory area.
As a Linux developer, I prefer this:
a. one box with 768 MB
b. one box with 64 GB
Janet Jackson's boobs!
JUST DON'T FORGET WHO HER BROTHER IS!!!
--
# Important Stuff: Please try to keep posts on topic.
# Try to reply to other people's comments instead of starting new threads.
# Read other people's messages before posting your own to avoid simply duplicating what has already been said.
# Use a clear subject that describes what your message is about.
# Offtopic, Inflammatory, Inappropriate, Illegal, or Offensive comments might be moderated. (You can read everything, even moderated posts, by adjusting your threshold on the User Preferences Page)
# If you want replies to your comments sent to you, consider logging in or creating an account.
Why not include Windows and perhaps others? I guess they wouldn't include non-open-source ones because it's a site about open source but I'd love to see the comparison. Have any other sites done that?
Comprison of various Linux distros (and of the 2.4 versus 2.6 kernel) is interesting. However, what is really lacking is an objective comparison of MS Windows Server 2003 versus Linux. I know Microsoft tries to prevent such benchmarking, but can they really enforce such a ban? It ought to be possible to find a team of Windows experts to tune Windows so the comparison is fair. Why not?
This is clearly a plot by Microsoft and SCO to destroy OSS! They couldn't beat any one of us, but if they get us to fight each other...
Personally I would like to see a corewar challenge where each OS team gets to write
apps that crash/dos/disable the others OS.
Last OS standing wins.
/me laffs
Like most men, I have been actively engaged in the Beaver Challenge since I was about 13 years old. I'm sad to say that other than a few phone numbers and some awkward dinner conversations, I haven't made any significant progress...
I have to confess to being more interested in the universality of the testing framework than any of the results. Whatever is done I hope it leads to some standards for future, lower profile but perhaps more useful benchmarks.
An accepted cross distro testing criteria would be nice.
ls
Dude, I know you're not supposed to read the story, but at least read the text you quoted.
This list is not final and if people want to ante in to try this with their favorite distro, let us know at bc2004 at osuosl dot org or in #beaverchallenge on the Freenode.net IRC network.
Note what's implied...
Please help metamoderate.
Gentoo will not be able to exercise any full potential it may have. If the people at the OSL install Gentoo it would be different than if the Gentoo team installed it with fully customized stage1, etc. Just an interesting though.
Come on, if it's built by Dell in 3min the
PS/2 & USB controller will die, 3hrs the HDD controller will fail & wipe each HDD, but in 3 days they'll replace them with the equivelant Compaq or clone machine....
Flamebait I know... Am I Ashamed? No
: )
Yet another crippling blow has struck what's left of the *BSD community, as a soon-to-be-released report by an independent commission doing a year-long study concludes: *BSD is dead and mummified. Here are some of the commission's findings:
.005% of internet servers. "It's just not reliable," said Christine McGee, VP of Technology for eBay, Inc. "Nor do we find it a very modern OS. I would recommend Linux to anyone contemplating a server OS, or maybe Windows, before I would recommend a BSD."
Fact: the *BSDs have balkanized yet again. There are now no less than twelve separate, competing *BSD projects, each of which has introduced fundamental incompatibilities with the other *BSDs, and frequently with Unix standards. Average number of developers in each project: fewer than five. Average number of users per project: there are no definitive numbers, but reports show that all projects are on the decline.
Fact: Apple is quietly changing the base kernel for OS X from *BSD to Linux. Insiders report that Apple's technical leadership has grown tired of the licensing battles and is seeking a more modern license; they find Linux's license more appealing. Many Apple technology experts -- from OS developers all the way up to Steve Jobs -- find Linux to be a more advanced OS, which will enable Apple to release a more mature product. The frequent hallway arguments and fistfights among the *BSD developers Apple has hired has also contributed to the decision.
Fact: XFree86 is dropping support for *BSD. The remaining core group believes that the *BSDs have strayed too far from Unix standards and have become too difficult to support along with Linux and Solaris x86. "It's too much trouble," said one anonymous developer. "If they want to make their own standards, let them doing the porting for us."
Fact: Many user-level applications will no longer work under *BSD, and no one is working to change this. The GIMP, a Photoshop-like application, has not worked at all under *BSD since version 1.1 (sorry, too much trouble for such a small base, developers have said). OpenOffice, a Microsoft Office clone, has never worked under *BSD and never will. ("Why would we bother?" said developer Steven Andrews, an OpenOffice team lead.)
Fact: servers running OpenBSD, which claims to focus on security, are frequently compromised. According to Jim Markham, editor of the online security forum SecurityWatch, the few OpenBSD servers that exist on the internet have become a joke among the hacker community. "They make a game out of it," he says. "(OpenBSD leader) Theo [de Raadt] will scramble to make a new patch to fix one problem, and they've already compromised a bunch of boxes with a different exploit."
Fact: NetBSD, which claims to focus on portability (whatever that is supposed to mean), is slow, and cannot take advantage of multiple CPUs. "That about drove the last nail in the coffin for BSD use here," said Michael Curry, CTO of Amazon.com. "We took our NetBSD boxes out to the backyard and shot them in the head. We're much happier running Linux."
Fact: There are almost no FreeBSD developers left, and its use, according to Netcraft, is down to a sadly crippled
Fact: DragonflyBSD, yet another offshoot of the beleaguered FreeBSD "project", is already collapsing under the weight of internal power struggles and in-fighting. "They haven't done a single decent release," notes Mark Baron, an industry watcher and columnist. "Their mailing lists read like an online version of a Jerry Springer episode, complete with food fights, swearing, name-calling, and chair-throwing." Netcraft has reported that DragonflyBSD is run on exactly 0% of internet servers.
With these incontroverible facts staring (what's left of) the *BSD community in the face, they can only draw one conclusion: *BSD is dead and mummified.
This is boring! It's all BSD and Linux! It would be really interesting to see how some of the new and completely from scratch open source stuff does too.
-Ted
-=-=- Quantum physics - the dreams stuff are made of.
They have failed to say which RAID controller is installed in the PowerEdge 2650, but I'm assuming it's the default Adaptec ROMB PERC 3/Di card. Following development of this driver on linux, there are issues with Linux and this driver. While I'm for a fair benchmark, this will most likely effect the Linux results.
A controller card agreeable to all OSs/Distros would be a good idea (if such a thing is even possible).
There are lies, damn lies, and benchmarks. I'm sure different configurations would produce similar complaints from other OSs/Distros.
I am, and always will be, an idiot. Karma: Coma (mostly effected by
What's with the x86- and Linux-centric approach? Do we really need to see how 6 different distributions can be tweaked to behave like one another on the same $4300 piece of hardware? I'd be extremely interested to see a G5 Xserve entered into that mix, although you'd clearly have to add some unnecessary doo-dads to the Mac to bring the price over $4000 (even with hardware RAID and the inability to drop below an 80GB HD to the 18GB like the Dell has, I could only bring a single processor Xserve up to $3500). Include a PPC Linux or two while you're at it. As it stands, the results will probably be at least a 6-way yawn-fest.
"In this competition they will be allowing a community of experts in each OS to tweak their configurations to ensure maximum performance."
I can see it now, teams of KDE and GNOME developers going head to head to see who can come up with the best color scheme, antialiased fonts, and 'Are you sure you want to delete this?' dialog box. Followed by Round 2, where each group has to compile something built for the other camp's desktop, whoever can fight through the dependencies quicker wins!
Lord Linus, save us from OSNews.
I want to delete my account but Slashdot doesn't allow it.
That's what I'm waiting for. I'm a champ!
there is no osnews.
that site never existed.
go away.
msn has better OS news.
that f00gly lady is scarey.
Funny how its mentioned here less than a day after I last watched it. Best sketch from the movie:
Hedley Lamar: Qualifications?
Hired Thug: Rape, murder, arson, and rape.
Hedley Lamar: You said rape twice.
Hired Thug: I like rape!
(everyone laughs)
And to stay on topic:
Beavers?! We don't need no stinking beavers!
Define "friendly." FreeBSD has some of the best documentation I've seen anywhere, a more than supportive community and has the same GUIs that Linux does. Once you've set it up (the installer is curses based, but makes a LOT more sense than most point&click Linux installers), it's no more difficult to use than any Linux distribution. Come on, you've got more sense than to say this.
The World is Yours.
just imagine... well you know.
What?
Just imagine what?
A Beowulf cluster of Beavers?
Will they let you tweak Hyper Threading?
It'll be interesting to see how many people turn Hyper Threading OFF when doing some tests. I found that my database was 212% FASTER for read operations after I turned Hyper Threading off on the 2650.
windows is not included, i would love to see how well linux and windows go head to head when properly configured. A win vs lin photoshop challenge, now that would be interesting..oh wait..
The war with islam is a war on the beast
The war on terror is a war for peace
I know benchmarks are cool and all, but I'd much rather see which distro performed well on older machines, like Pentiums and Pentium IIs. I have so many of these to deal with, and most everything sucks. Funny, cuz I remember running Solaris x86 and OS/2 and thinking they were excellent (sort of). Now, everything seems bloated. Of course, I'm expecting too much...
Even more so if the Gentoo people manage to compile their stuff with the Intel compiler. Everyone I've talked to says they see speedups of around 100% when apps are compiled with that thing.
I don't know, however, if it's possible to emerge Gentoo apps with icc instead of gcc. I'm using Gentoo on PPC, so it'd be hard for me to test
one hundred twenty
is just enough characters
to write a haiku
This contest is blatantly set up so that Gentoo, the obvious victor in a fair benchmarking contest, cannot win.
They are only allowing three days to set up the OS, everyone knows that you can't get gentoo installed, much less customized in that time.
And it was 523% cheaper, took 45 days less time, and tasted like chocolate.
Karma: Bad (mostly due to all those "In Soviet Russia" jokes)
I wonder when they will bentchmark the install times, like for gentoo it is extreemly slooooow.
[blue] - The Ministry of Information approved this message...
I find a general lack of a definitive hardware list. I presume this can be had by subscribing the to forum or mailing list, but I really think this should be published up front in a no nonsense manner.
Or maybe I missed the page with the list... It happens.
I think you underestimate just how much I just dont care.
So let me get this strait.
N GNU/Linux distributions are going to be
compared to 1 each of the other major OS's.
So the best results for a particular benchmark
are going to be selected and compared with those
from the non Linux OS's.
That doesn't sound like a valid comparison
method.
Not to nitpick or anything, but the phrase "very unique" is a pet-peeve of mine. It's a (futile, imho) attempt by the ineloquent to underscore the originality of something, resulting only in them looking stupid. Because this is Slashdot, and I'm such a nice guy, I'll let you in on this so that you don't look idiotic yourself in the future.
Unique by definition means that there is only one of whatever is being called unique; that's where the latin prefix uni- (meaning one) comes into play, semantically. Therefore, there cannot be degrees of uniqueness -- by definition it is a boolean state, being either unique, or, well, not.
So, in sum, nothing is very unique.
It's mindlessly redundant.
Thank you, drive through.
For me, apt-get really shines when ~removing~ packages, moreso than when installing them. That's still its prime advantage over up2date, last I used up2date.
I've been curious about Gentoo; does portage offer the removal advantages that apt-get does? I.e. when you remove a package, will portage remove all the packages that depend on that package, too, saving you the hassle of removal-dependency-hell?
Andrew Klaassen
Is there any benchmark comparison of these OSes on laptops? It would be really interesting to see how they manage APM/ACPI/... to balance performance and battery life.
"Open source" has taken on a whole new meaning since Hustler got involved...
bytesmythe
Hypocrisy is the resin that holds the plywood of society together.
-- Scott Meyer
That gay perv hemos.
2 Procs - comodity, lets see how the various distros do on 16-64 way servers
2 GB RAM ? why so limiting - lets get this up to 64 GB - or more
The disk system might not be too bad - but hardware Raid 5 would be more realistic
Of course what they are going to do is figure out which distro runs a FPS with the highest frame rate
I have mod points and I am not afraid to use them
If one of the reasons to run the benchmark is to show people how great F/OSS is, "round 2" seems like a natural. After all, don't we all know some people who have a couple of heavy paperweights sitting around? Wouldn't more people be willing to try something other than Windows if they saw that their old crappy box ran just as fast as their new shiny WinXP box?
Yet another crippling blow has struck what's left of the *BSD community, as a soon-to-be-released report by an independent commission doing a year-long study concludes: *BSD is dead and mummified.
.02% and is sinking further. Netcraft comments that the constant in-fighting among *BSD developers is probably to blame.
.005% of internet servers. "It's just not reliable," said Christine McGee, VP of Technology for eBay, Inc. "Nor do we find it a very modern OS. I would recommend Linux to anyone contemplating a server OS, or maybe Windows, before I would recommend a BSD."
Here are some of the commission's findings:
Fact: the *BSDs have balkanized yet again. There are now no less than twelve separate, competing *BSD projects, each of which has introduced fundamental incompatibilities with the other *BSDs, and frequently with Unix standards. Average number of developers in each project: fewer than five. Average number of users per project: there are no definitive numbers, but reports show that all the projects are on the decline, because *BSD is dying.
Fact: Apple is quietly changing the base kernel for OS X from *BSD to Linux. Insiders report that Apple's technical leadership has grown tired of the licensing battles and is seeking a more modern license; they find Linux's license more appealing. Many Apple technology experts -- from OS developers all the way up to Steve Jobs -- find Linux to be a more advanced OS, which will enable Apple to release a more mature product. The frequent hallway arguments and fistfights among the *BSD developers Apple has hired has also contributed to the decision.
Fact: XFree86 is dropping support for *BSD. The remaining core group believes that the *BSDs have strayed too far from Unix standards and have become too difficult to support along with Linux and Solaris x86. "It's too much trouble," said one anonymous developer. "If they want to make their own standards, let them doing the porting for us."
Fact: Many user-level applications will no longer work under *BSD, and no one is working to change this. The GIMP, a Photoshop-like application, has not worked at all under *BSD since version 1.1 (sorry, too much trouble for such a small base, developers have said). OpenOffice, a Microsoft Office clone, has never worked under *BSD and never will. ("Why would we bother?" said developer Steven Andrews, an OpenOffice team lead.) With fewer and fewer apps available, it is only a matter of time before *BSD dies and is mummified.
Fact: According to a brand new Netcraft survey yet to be published, the number of internet servers running *BSD has shrunk to a heartbreaking
Fact: servers runningOpenBSD, which claims to focus on security, are frequently compromised. According to Jim Markham, editor of the online security forum SecurityWatch, the few OpenBSD servers that exist on the internet have become a joke among the hacker community. "They make a game out of it," he says. "(OpenBSD leader) Theo [de Raadt] will scramble to make a new patch to fix one problem, and they've already compromised a bunch of boxes with a different exploit."
Fact: NetBSD, which claims to focus on portability (whatever that is supposed to mean), is slow and cannot take advantage of multiple CPUs. "That about drove the last nail in the coffin for BSD use here," said Michael Curry, CTO of Amazon.com. "We took our NetBSD boxes out to the backyard and shot them in the head. We're much happier running Linux."
Fact: There are almost no FreeBSD developers left, and its use, according to Netcraft, is down to a sadly crippled
Fact: DragonflyBSD, yet another offshoot of the beleaguered FreeBSD "project", is already collapsing under the weight of internal power struggles and in-fighting. "They haven't done a single decent release," notes Mark Baron, an industry watcher and columnist. "Their mailing lists read like an online version of a Jerry Springer episode, complete with food fights, swearing, name-calling, and chair-throwing." Netcraft has reported that DragonflyBSD is run on exactly 0% of internet servers.
With these incontroverible facts staring (what's left of) the *BSD community in the face, they can only draw one conclusion: *BSD is dead and mummified.
Typical Slashdot Response #1 = Take bets on the winning score
Typical Slashdot Response #2 = "all of us
Typical Slashdot Response #3 = Being
(those few born with one are naturally excluded from the competition, as they've automatically scored 1 even before the competition starts, which by the letter of the rules means they cheated)
Typical Slashdot Response #4 - Anyhow, they prefer to be called Vagina Squirrels
I did a search on google, and found these rules for Beaver Challenge 2004 (PDF).
Visit CryptoGnome in his home.
I guess this would involve som serious tweaking...
A cheapo Dell blade is "wicked"? Dude, you have no clue whatsoever.
...well, at least not if you ask a *BSD user.
Great! Where do i sign up!!!??!??! :P~~~~~~~
I am soooooo sick of people hyping this distro.
I spent weeks of my life getting it working on a P4 just to test out the claims.
Result?
No appreciable difference to any other distro with simillar configuration and a LOT more work.
Use what you want just please stop the teenage exageration.
Eugenia censors opinions to suit her amateurish pro monopoly$oft agenda.
I am soooooo sick of people hyping this distro.
I spent weeks of my life getting it working on a P4 just to test out the claims.
Result?
No appreciable difference to any other distro with simillar configuration and a LOT more work.
Use what you want - just please stop the teenage exageration.
Bogomips
'nuff said.
The final list isn't complete, but why isn't QNX mentioned %)
No "Um's" please.
So?
Just cause you thought it was a hassle to setup (which it might be, I dont know since I havent tried it yet) why is it "teenage exageration" to suggest that it should be in the benchmarking too?
Seriously, all he said was "isnt there a free beos", not "OMFG DUDES! There's a free BeOS out now and it pwnz all j00r linux and BSD boxes with its almighty powers of doom! Go get it now! And the reason it isnt in the benchmark is that it would k1ck j00r azzez the instant it touched the computer." Now -that- would be teenage exageration.
And, seeing as I managed to not read the topic of your post before I went off and posted myself, I misinterpreted what you said.
First of all, Gentoo is good.
Second of all, Gentoo isn't the easiest thing to setup.
Third of all, it is a matter of taste. And just as we shouldnt judge the general users of GNU/Linux on the mad zealots that shouts the loudest, you shouldnt judge Gentoo on the Gentoo zealots.
I use gentoo and I love it. On the other hand, Ive also used pretty much any other distro out there. IMHO, the one that was most pain-like to set up was debian. Though it is, a matter of opinion.
Now, I will go hide under a rock for a month.
What seems to be lost on the /. crowd is the inherent human variable that will bias the results. Whose experts are "better"? That may be more pertinent on benchmarks than the OS tweaks. Some experts will pick the "best" combination of optimizations, others will pick "suboptimal" combos. Two OSes that are "theoretically" equivalent are not likely to perform as such, because of this human variable. There will likely be a false winner, too.
OSNews should build one very large team of experts, that plan and configure all the setups concurrently.
``Afterall, they're all going to use the latest 2.4 and 2.6 kernel and comparable glibc versions''
How about a 2.6 kernel with statically linked busybox/dietlibc userland? Could run pretty fast...
But, as somebody asked, what are they measuring?
Please correct me if I got my facts wrong.
"just imagine... well you know."
A beowulf cluster of linux idiots. No, I can't. Oh, wait a minute ---- I've got it. Slashdot.
I wouldn't be all that sure about your predictions.
There are many things we don't know yet:
- What is going to be measured?
- How are measurements going to be weighted to compute the final score?
- What systems are participating?
The list of participating systems is not final yet. There are other systems out there besides *BSD and GNU/Linux. I could imagine an embedded Linux (without the weigt of a full GNU userland) beating the other Linuxen.
Certain things are pretty inefficient under UNIX and like systems (e.g. monitoring file descriptors for incoming data - although the *BSDs have kqueue and Linux 2.6 has epoll, see this benchmark). A non-UNIX contender might join and beat the others.
Hardware compatibility seems to vary among Linux distros. This could also affect scores (e.g. lost time getting things to work, that could have been spent on tweaking for performance).
Please correct me if I got my facts wrong.
So we now have two good reasons for doing without Hyper Threading: reduced cache size per virtual CPU, and increased complexity of the scheduler.
I have always wondered what the rationale for HT was. Clearly, it's increased performance for multithreading, but how?
The reasons I can come up with are reducing the number of context switches, and effectively increasing the number of registers. But then I look at a proper CPU (Alpha, MIPS, PPC, take your pick) and I see vastly more registers, and I hear that context switches are far less expesive there as well. All this wihout splitting cache and the extra complexity in the scheduler.
What am I missing?
Please correct me if I got my facts wrong.
I've been stuck on older systems more often than not and as a result I spend time tweeking the box to make it more effecent.
I memerise all my tweeks and improve on them as time progresses.
I know there are diffrences beween optomising on the low end and opomising on the high end.
But I think I know enough that I can do it.
I don't actually exist.
I'm seeing 279 days of uptime on both my 2650s running a very minimal install of RH7.2 with some (not all) errata packages applied, kernel is 2.4.18-24.8.0smp, a tad behind the bleeding edge, but serviceable nontheless.
They both have 100+ Gb RAID5 on Perc3/Di controllers, and are running a reasonably demanding application (Apache/Jakarta/Servlet, MySQL, Verity K2 with ~20Gb in collections so far..)
from the build documentation that I wrote 281 days ago:
*
* WARNING - RedHat 7.2 does not autodetect
* Perc3/Di RAID controllers correctly
*
- enter 'expert noprobe' at RedHatboot prompt
- say "No" when asked "Do you have a driver disk"
- choose a language, keyboard type, and installation method
- select "Add Device" when you get to "Devices" section
- select "SCSI" as "Device" type
- move down cursor to "Adaptec AACRAID (aacraid)", choose the "Specify module parameters" box
- enter aacraid_pciid=0x1028,0x0A,0x1028,0x011B in the dialog box
- select done
$ strings FTP.EXE | grep Copyright
@(#) Copyright (c) 1983 The Regents of the University of California.
I haven't looked into performance tuning of machines, but off the top of my head I have installed ATLAS (Automatically Tuned Linear Algebra System) on a few things. Would the results of installing ATLAS on a computer be any help in setting general compile flags for all the other things which are installed? Are the processes used by ATLAS useful for tuning other things?
Gentoo is the coolest because it boots to a kick-ass purple framebuffer terminal.
If they can't get the 2650 to work, I'll donate my toaster-oven.
I am the Lorvax, I speak for the machines.
Sorry to quibble, but "A musician needs the RIAA, is like a fish needs a bicycle" simply sounds better. I agree entirely with the sentiment, however!
~Morosoph
This!
Okay, there's no "beaverbeaverbeaver.com", but there should be!
~Morosoph
The first version I recall is 5.1 was which indeed based on Red Hat. From the Mandrake 5.1 release notes.
"Linux-Mandrake is an updated Linux-RH 5.1 GPL, with KDE 1.0 fully integrated and preconfigured in it. Those two parts have been (not so much) modified and improved to work properly together."
Red Hat didn't use to include KDE so we had to download it and configure it ourselves. Actually it was pretty easy. But for those who had heard of Linux and wanted the easiest to use DE (because of KDE) many tried Mandrake. I actually was a Huge Fan of Mandrake till about 7.2 IIRC. Man was that a good release. But since then I've simply not had anything good to say about their QA. Not to flame but IMHO since they simply stopped "building on" RH releases their quality has gone way downhill.
If you wanna get rich, you know that payback is a bitch
export CFLAGS="-Wall -pipe -O2 -fomit-frame-pointer -fno-strict-aliasing \
-mpreferred-stack-boundary=4 -malign-double -march=athlon-xp -frename-registers \
-fprefetch-loop-arrays -falign-functions=64 -ffast-math -mno-align-stringops \
-mfpmath=387 -fnew-ra"
export LDFLAGS="-Xlinker -s"
Today, don't use the options -fssa and -fschedule-insns (and related of the same name) because it crashes.
I don't use -maccumulate-outgoing-args -finline-functions -funroll-loops because it generates FAT CODE !!!.
open4free
For kernels, i should use the option -mpreferred-stack-boundary=2 [2^2 = 4 bytes] (instead of -mpreferred-stack-boundary=4 [2^4 = 16 bytes]).
open4free
At the very least its open tools on an NT kernel (assuming cygwin on 2k or xp or other stable windows platform)
OpenBSD being in this competition is silly. I say this as a user and lover of OpenBSD since 2.5.
OpenBSD focuses on extreme security, which goes beyond just auditing code. It has mechanisms in place, which are specifically there to heighten security first and foremost. Quite happily foregoing speed and scalability (in places where it is unavoidable).
Examples are, consistency checking all over the place and randomisations, which provide greater security but inevitably, hurt performance.
This would be like comparing high performance cars and throwing in an M1 Abrams main battle tank to also be tested as a high performance car. In the end, the tank will obviously have terrible top speed, acceleration and cornering ability and everyone will pronounce it crap. But try defending a nation with Ferrari 360 Modena's.
This course is not for this horse. What's more, OpenBSD has been very tentative about moving forward on SMP for very good reasons. It opens up a huge can of worms (bugs), which makes understanding code and thus auditing extremely difficult. The machine being used for the test is SMP!
FreeBSD SMP is supposed to be fantastic in the new technology releases and so-so in the stable releases.
This test would be better conducted in about a year and without OpenBSD (even if it did have SMP by then). OpenBSD is never going to be the highest performance server because that is not what the OpenBSD project strives for and they are willing to hurt that area in the name of security.
If this machine were not SMP, my money would be on NetBSD with FreeBSD coming a close second.
War crimes, torture, lies, illegal spying... Would someone give Bush a blowjob, already, so he can be impeached?
Exactly like that, but folks have short memories, and Democrats can do no evil in the eyes of fellow Democrats.
Like the old saying goes, power corrupts.
For all of 11 months of 1993 and all of 1994, Clinton had a democrat controlled House and Senate. They got to push through anything they wanted.
Now that the shoe is on the other foot, Bush has a Republican controlled House and Senate all the democrats can do is complain about it.
I'm a Republican. I will be voting for Bush in November, not because I think he's perfect. In fact he is flawed in many ways, but I'd prefer him over any of the Democrats who have a chance to win.
LK
"Hi. This is my friend, Jack Shit, and you don't know him." - Lord Kano
that's not a troll, a troll is +4 mod on a content-free post...