Slashdot Mirror
madddog on Linux v NT Benchmarking
BogoMips sent us an interesting tidbit running in Performance Computing currently. Jon "maddog" Hall explains some of the benchmarking issues associates with the DH Brown reports, as well as the ubiquitous Mindcraft tests. Very well written article, IMHO.
But don't whine here when Red Hat "downsizes" and Mandrake has nothing to take credit for.
If M$ and other commercial software vendors were more forthcoming with known bugs/issues, those per incident calls would be needed less often.
Case in point: While trying to get the Exchange agent for ArcserveIT installed, I was having some difficulties, I wasted a few hours with it. then checked their webpage... nothin. So I called CAI (ugh). The response I got was: "oh we broke that, ya gotta do this....."
Sometimes ya gotta pay the ba$tards to get any answers.
IMO thats one of the best features of Linux, you can get free support (via LDP, IRC, mailing lists....) You would be hard pressed to find the kind of grassroots support that linux has, in the NT/M$ world. Broken features are usually well documented in most OSS apps.
See, this is exactly what I'm talking about. Most of what you wrote above is way over my head. As simple as it sounds to you, even for experienced IS managers (granted, who's entire world revolves around Microsoft) are clueless when it comes to recompiling a kernel.
I do agree with your position with regards to grabbing system information, and optimizing the kernel itself. That would kick serious bootyack.
"See, this is exactly what I'm talking about. Most of what you wrote above is way over my head. As simple as it sounds to you, even for experienced IS managers (granted, who's entire world revolves around Microsoft) are clueless when it comes to recompiling a kernel."
They shouldn't be touching the kernel, the sysadmins do that. My manager doesn't touch my servers for any reason either.
Chris
mtnbkr@mindspring.com
Neither NT, nor any Unix "kick ass" right out of the box. All require some effort in tuning to make them run efficiently for your needs. How many servers have you installed?
Chris
mtnbkr@mindspring.com
What kind of corporation requests a mail server be put up on a 486 w 16 Megs in this day and age?
We just throw money into systems around here.
Linux does have the advantage of being able to optimize for specific hardware, however the optimization really isn't that good. Compared to recent offerings from either Borland or Microsoft, gcc on the x86 does a very poor job optimizing code. I've heard that gcc does an even worse job optimizing for Sun and Alpha systems, but I can't confirm that as I only own x86 computers.
EGCS and PGCC are getting better at optimization, but they have a long way to go before they reach MS and Borland. It will be interesting to see if the new IA32 backend to egcs (supposedly appearing sometime after gcc 2.95) fixes some of these problems. I certainly wouldn't be surprised, given the pace developement has been on since cygnus started working on egcs. But for now, the default optimization you get with NT programs is alot better than what you'll be able to do with open source compilers.
Remember folks. John wrote this at the end of
April before the c't benchmarks came out.
Microcraft was not the last word on Linux vs NT
benchmarks. Don't get too worked up.
Most of what I want to say here is merely echoing what you've already seen.
Linux is very good at what it was intended to do; Linux makes obsolete hardware useful again. Linux is also surprisingly good at many things it has been adapted to do, such as dept. file & print services, web serving, mail serving, etc.
When Linus started this journey we're all a part of now, all he cared about was his 386. He had no idea that we would ever be doing what we're doing today with Linux. Scaling up to big iron and scaling down to the smallest web server in the world. There are new surprises every day.
Will Linux function as an enterprise server? Yes. Will it meet the expectations of this role as well as a platform intended for this use? Not today. This is not a loss.
When you can't take the latest distribution of your choosing, pop it on a 486 with 16MB of RAM, and have a dept server, now there is your loss.
Why doesn't MS use them on hotmail.com and homepages.msn.com? there are probably other sites too that use Apache. Look at http://homepages.msn.com/asdfqewr (an error page) for the bit that says Apache at the bottom.
Yes, that is exactly how I feel (although I think that Linux will keep doing well).
...
When Linux can beat AIX, I will use it. Until then, I will keep on with boxes like my s70s and my F50s.
Although I love it at home
Brendan
wars fought, bars emptied, insurance collected, AIX wrangled
What was your problem with Portland Group's compiler? I've used their Fortran compiler for some scientific code and found it to be excellent. When adjusted for the relative speed of the machine based on specfp numbers, I found it gave me comparable numbers to Sun's and SGI's compilers. I don't have any experience with the C compiler though. I'd have to agree with you on the Alpha. We looked into buying one and running linux, but I couldn't find anyone who had anything good to say about Microway's compiler. See http://www.pgroup.com.
I must say that watching this Mindcraft saga unfold has been great. It showed how fanatical Linux users can be. No amount of real data prevents them from throwing insults and accusing everyone of conspiring against them. When the first test results came out they were stunned, then there were tons of "if they just did that, the results would be the other way around" comments mixed with baseless attacks against everyone involved. Then there was hope when the retest was announced and even more hope when those idiots at Microsoft sent their marketing people against best Linux hackers. When Linux lost badly again, the amount of insults didn't change - only their targets did.
Take a look at yourselves people and think about how others could perceive you.
The computer is easy to get, the OS is hard to get.
If one were, say, to have a couple of, say, surplused 4.2 CDs (hypothetically, like 120 sets of the damned things), where would that someone (not me, of course) send them so that an interested poster such as yourself would get to you? Hint, hint. Always good to spread the knowledge, I say. But I am staying AC. Naturally, I am not saying 4.2, the Performance Toolbox CD, the Domino that comes with it, the Bonus Packs, the Netscape Navigator, the AIX Rocks CD with Doom for AIX, and so on, because I am not saying anything at all. Just suppose that someone else is saying this. No, you don't get the jewel cases, that someone needed them.
And yes, damnit, I like AIX. It takes enough flak from the SunOS weenies crying into their beer about being nonstandard. Hmmf! Let me see them move around 200GB of files WHILE THE USERS ARE WORKING WITH THEM during the working day from disk array to disk array without a hiccup or a minute of downtime! AIX can. I don't like to do stuff like that, but it can.
Now if the fscking 7017 power supplies would stop croaking on me, I would be a happy bofh.
I could'nt write a line of code to save my life.
/usr/src/blah, followed by make config.
But I was perfectly capable of rpm -i blahblah-src.rpm, followed by cd
I've installed NT. The process might be more aesthetically pleasing, but it's not really any harder, assuming the user is capable of reading the directions that are printed on the screen.
Pgcc might do wonders for pentium class chips but hardly does anything for P6 cores. Something else that is somewhat interesting is that rc5des cracks RC5 keys at the same rate on my new K6-2 300 that it does on my old PPro 180 (actually 150 oc'd to 180).
Also, I had compiled mysql for a production server with pgcc 2.91.66 (using -mcpu=pentiumpro -march=pentiumpro) and noticed after 10-20 days, it liked to suddenly die under heavy use. Recompile it with gcc and everything was fine. It is a dual P2-400 system so it could be a bug somewhere in the kernel, but all I know is it has been running fine for the last 45 days.
The LDP is god as far as I'm concerned.
But I've tried to get help on IRC a number of times (efnet and undernet), and met with nothing but hostility.
Hence, the LDP is god.
The cost of the OS only matters to lower end sites anyway. If you are going to spend 6 figures on hardware to serve millions of hits a day while streaming out videos right and left, the least expensive item is going to be your OS.
I think the slashdot community as a whole is being a little too hard on Mindcraft. Was the test unfair? Sure. Was their a concious premeditated decision to skew the results dramatically? I kind of doubt it, actually.
The test was done by NT people. Of course they're going to pick the best system for NT, that's what they're used to. Are they going to know how to make the same adjustments to linux that they would to NT? Probably not.
Microsoft has done some shady stuff with non-disclosure of their funding of "independent" reviewers, etc. in the past. But even in the first Mindcraft test they were perfectly upfront about saying where their money was coming from.
So the linux community (rightfully) gets agitated and points out the flaws in the test. The most fair thing mindcraft could possibly have done would be to redo the same benchmarks (OS's in an off the shelf high end commercial server) --inviting crews in to optimize "their" systems. Which they did.
The second mindcraft test might still have used a hardware setup we'd consider weird, but that's why other companies are doing DIFFERENT benchmarks with different hardware setups.
Most people just pissed and whined like the boy who cried FUD. Others sat down analytically, recgonizing suboptimal conditions in the mindcraft tests and designed spiffy new tests for linux's power to shine.
The second is preferable.
Of course hotmail.com doesn't run Linux either. [htomail.com does though]. I think the most objectionable aspect to these benchmark "tests" is they didn't test more operating systems! It is a joke to think the world revolves around NT and Linux.
I thought I read a few months back that Intel was going to help improve egcs' performance. I also read someplace (sorry, don't remember where) that Intel's own compiler was the best x86 compiler around (for optimizing code).
Does anyone know if this is true? It seems reasonable that Intel would have the expertise required to improve compilers for their chips. They also wouldn't mind seeing their products do on benchmarks under Linux.
Most people may have missed the fact that Mindcraft tested on US $50K equipment.
Realistically, at this price, most people would choose better price/performance of big boys like sgi/sun/ibm/hp, even small mainframes that put NT & Linux behind, anyway.
A better benchmark would be on today's realistic equipment.
Enterprise servers are essential to many companies, they need big iron to carry alot of tasks. And yes, it is possible at times to split these tasks across many smaller machines, but the operational and management issues tend to make it inefficient and down-right hard work. Which is why alot of companies are spending alot of money on server consolidation, bringing their distributed servers back into the centre. Be it into Lan-in-a-Can type boxes like the IBM SP, big SMP boxes like UE10ks etc. At the end of day it is just easier to manage.
Linux will get better at this, tools which allow you to efficient manage clustering, fail-over type scenarios etc,etc but it isn't there yet. There are some good foundations being built and there are plenty of examples to look at and learn from. For instance, someone might want to look at IBM's PSSP which is used to manage the SP environment, there is alot of stuff implemented in Perl and TCL/tk there. IBM has been using open-source products for a long time in a vey real way, just not many people realise.
Slashdot DOES NOT run on FreeBSD. They are playing with FreeBSD to display the banner at the top for adfu (or they were). The ad-banner still shows up unreliably so what does that say for FreeBSD?
Its urban myth that FreeBSD is the be-all-end-all of networking. Ya it runs on Yahoo but it is not stock FreeBSD out of the box. Yahoo is not very impressive.
Deja.com runs Linux and it certainly is doing more work considering the volume of usenet news articles appearing daily.
At least I can stand using Linux, as opposed to AIX :)
I attended the panel discussion at the AIIM show
that Maddog is mentioning in his article. Something funny happened:
- moderator (to M$ guy): There are rumors that MS is porting Office to Linux, is that true?
- M$ guy: absolutely not.
- moderator: will you in the future?
- M$ guy: no, never.
- moderator: why?
- M$ guy: because there is no demand for it. Our customers aren't asking for it.
- moderator: let's ask the audience...
- M$ guy: how many people use Linux on a regular basis as their main computing platform?
[about 100 hands are raised in an audience of about 1000, the other 900 cheer the 100]
- moderator: how many people want M$ to port Office to Linux?
[80% of the audience raise their hand. The audience laughs and applaudes loudly].
It sounded like many of these folks would switch to Linux if it weren't for MS Office.
- Anonicous Moward
If /. DID run on FreeBSD, that'd be a good reason to steer clear of it. /. is one of the slowest sites on the Net, period. The performance is absolutely laughable.
Blah, blah, my OS is better than yours blah, blah, blah....
First, remember Linux Advocacy shouldn't be mean-spirited. That said...
Since I've been maintaining a mission critical app (January 2, 1998) I've been paged exactly zero times. This is due, in large part, to quality development work, tested hardware, and backup hardware. All of these were available quickly, and inexpensively.
One of my colleague's systems, Solaris-based with Oracle on the back-end, is just about as reliable. However, he paid out the yang for developers that knew their stuff around Solaris and Oracle. He also needed to find supported hardware. The best he could find was a contractor that gave him a 24 hour response window, and ultra expensive (by a factor of 5) hardware. Meanwhile, any errors that occur in his system take far longer to diagnose and repair, and end up costing tons more, than in my system; all due to the limited availability, and high cost, of support and hardware.
TCO, keep in mind, also takes into account the availability of hardware and software support when (if) it's needed. Sure, when it's done right, you shouldn't need the support, but when you do, expect to paaaaay.
Believe me, I am a fan of Linux. I want it to succeed. The very least it can do is force Microsoft to adjust to stay competitive, which makes my life tons easier.
This is why I'm arguing against this topic. I don't want Linux Advocates resting on their laurels by saying "Well, code optimization is the answer." No, it's not. Linux should be smarter than that out of the box. When it is, it'll be another chink in MS's armor. THIS IS A GOOD THING.
Done. Enjoy. Make more money. Help someone else.
yeah, like a benchmark study done by the linux people would be impartial. I can read the headlines already "Linux found 10000 times faster than NT". I wouldnt trust it either.
You guys are as bad as M$ when it comes to making false claims.
The problem with this line of thinking is that not everyone is a kick-ass coder, and isn't up to the challenge of this kind of optimization. It's like saying that "The General Motors XF-3000 goes from zero to 60 in 1.3 seconds, and gets 80 miles to the gallon, but you have to optimize the frame and engine for the size, and weight of the driver, and for the road conditions.
Luckily, Windows NT Server kicks ass right out of the box. Neat, huh?
I would look on ebay. You can often find old ppc RS6000s for $500 or less. Get one, get a tape, get a CD, and get an old terminal and you will be in business. It will probably come with the OS. Back that up to tape right away -- and save that tape! Otherwise it will cost you bucks. Patch up from IBM's web site. Enjoy.
If you can hork a copy of AIX from someplace, even better. Just in case.
AIX is odd, it is IBM, but is is very easy to work with in obsenely large and loaded enviroments.
That is what I actually did, BTW, although I haunted Rice and Univerity of Houston auctions for the boxes, as ebay wasn't around a few years back!
Last month, the German Magazine decided to check
the results of the last MindCraf Benchmark.
However, they decided to test performance on
many hardware configurations.
The results were this:
1CPU + 1 NIC --> Linux wins.
Several CPUs + 1 NIX -> Linux wins.
1 CPU + Several NICS -> Linux wins.
Serveral CPUs + Several NICs -> NT Wins.
99% of servers don't use only one NIC (with one or many CPUs).
Conclusion: Linux is faster than NT in 99% of the hardware configurations.
Meanwhile, Linus already told that he is addressing the botleneck caused by the "global lock" in the networking sub-system, that caused this problems in the MindCraft Benchmark.
The solution (Linux 2.4) is expect to the end of the year.
Yes, we are talking about the little free Linux, agains't the expensive proprietary NT giant...
The problem with linux and large enterprise servers is mostly in it's style of development. Most people don't have access to an Ultra Enterprise 10000 ``starfire'' and several disk arrays to just sit around and play with -- they are stuck with (relatively) cheap PC's and Suns and Alphas. Only large companies can afford to shell out millions to buy the equipment, and millions more to pay programmers to develop for it. That's why Sun Enterprise equipment is almost mission-critical-environment ready -- and linux isn't. You can't yank a processor board out of a machine running Linux and still have the thing hum along. Yet.
-- Erich
Slashdot reader since 1997
Of course, most of the "enterprise" settings I've worked in have featured NT with "admins" who couldn't find their ass with both hands. It always amazes me that companies are willing to pay big bucks to people who, whenever a problem comes up, just go running off to a pay-per-incident help 800-number anyway...
----
Every year during my review, I just pray the words "slashdot.org" aren't mentioned.
----
Every year during my review, I just pray the words "slashdot.org" aren't mentioned.
Those people who want to rant and rave when anybody says anything slightly bad about Linux should read this article.
...phil
...phil
"For a list of the ways which technology has failed to improve our quality of life, press 3."
I have enough diskspace to hold source for most of the Linux stuff I have. I want to compile it from scratch (I'm quite prepared to). Debian does not *need* to do what you are suggesting, if they fold source support properly into apt, as they've been talking about for long enough, then they will be able to support everybody, as those like us can simply do "make world" and rebuild the entire system, or even, with some work, allow all the systems to be easily supported - without the hassle of fiddling with code
The issue that the article was talking about was whether Linux was suitable for very high-end servers. High-end servers aren't the domain of NT, but rather the commercial Unices put out by the likes of Sun, IBM, HP, and so on.
This is pretty much an issue of Unix flavor vs. Unix flavor. Linux just happens to still be a fairly low-end Unix flavor--at least for now.
Posted by MaverickPl:
What is it??
? I've tested this, and gcc is pretty good, and egcs is great on x86. I remember that when comparing it to the old Watcom C compiler (which was the best back then) integer performance was always equal or up to 30% better, and only floating point performance was lacking by a percent or two. I wouldn't mind having a faster compiler, but what we have is pretty good, and is supported on many platforms. For really speed-critical optimizations, assembler still works.
pb Reply or e-mail; don't vaguely moderate.
I agree, I love it when I see benchmarking like this. tomshardware does a good job for this sort of thing with processors, but that's most of what they benchmark.
:)
I benchmarked emulator overhead for a while, but gave up when I realized that the actual processor overhead by running DOSEmu is a percent or two, and just completely installed Linux instead.
(recompiling benchmarks is also a good way to test compilers...
pb Reply or e-mail; don't vaguely moderate.
Recent versions of pgcc and binutils have support for K6-2,3 specific things, 3dnow for example. If you're whining about the K6(not the K6-2,3), its pretty much a P5, so you can still see a performance gain over i386 binaries if you compile with pgcc.
Now, to install almost anything, and have it completely optimized for my system, I go into /usr/ports/, pick the application, and do a "make install distclean" and it's installed seemlessly and optimized for my K6 :-)
I was refering to _every_ level of software, I suppose Apache would be the most obvious example with all the httpd comparisons going on lately. It's not just the kernel you can optimize.
And you don't need to make the changes instantly, you can do make, and then later when your ready (after a backup, and work has stopped for the night), do the "make install."
If your into the Linux "kernel of the week" syndrome, well, optimizing or not doesn't make any differance in how often you reboot, only the speed after the reboot.
If that is true, Intel could make a great marketing move. They could recover a lot of the "Anti-WinTel" fallout where people are moveing to Linux/AMD by open-sourcing thier compiler and helping merge it into egcs. It's like $700 now though? But, if you could make absolutely outragesly fast binaries for Intel CPU's out of normal open source apps, they would gain a lot of support in the Open Source community.
After that, to install, all you need to know is three words, "make install distclean", and even that third word is optional.
It does bring up the point that it should and could be easier. What is really needed is an expantion on the basic "uname" call, to include much more system information (specifically, the CPU, amount of memory, etc..) Then, it would be possable for open souce compilers to just make it one generic flag (like SGI's compiler does, -Ofast) and it would grab the system information for you and figure out the best flags on it's own. That is a realistic possability, and if something like POSIX, the LSB, or some other standardization body would implement this type of system call to get hardware information, it could potentially benifit the UNIX community in a way that Microsoft can't keep up with.
System tuneing is IMPORTANT. Important enought that it can make one OS faster than another. I think we should be pointing out that open source is much more tunable, not only in the ability to modify the code, but also in the ability to optimize it for specific hardware. (quick note on some stuff I tried to see the diffrence, click here)
Why can't someone do some intellegen testing on this, and give credit to the people who REALLY make GNU/GPL and all of open source a success, the folks who write the COMPILERS! Linus did write some nice stuff, as well as many others, but without the right compiler, it's worthless.
If I can do the work of a big enterprise server on four P75's, that's still a savings. Linux is proving that for some tasks, a cluster of small machines works just as well as one huge machine. And there's reliability projects in the works as well based on the same principle: have an extra cheap machine which grabs the load if your main server goes down.
These are examples from a large bank
IBM parallel sysplex running OS/390
IBM SP frame running AIX
Sun Starfire running Solaris
HP V2500 running HP-UX
NT doesn't even come to the knees of one of these systems.
pssst buddy... he quit Compaq already :)
jon@valinux.com iirc
Various ramblings
Agreed, however after reading it twice they still wouldn't understand.
Yes my OS doesn't have the fastest GUI, and it wasn't coded by 4000 paid employee's, and hell it might not be the best... but you know what? I didn't pay a damn thing for it. So who cares.
If I drove around a station wagon with a cracked windshield, that I got for free... I wouldn't bitch one bit. I would keep driving to work with the biggest damn smile on my face. Just laughing on the inside to all those fools who had to pay for their cars.
Awesome!
Jon did mention what is needed: the ability for the system to say up and available 24x7 REGARDLESS of disk failures, CPU deaths, and motherboards frying.
Linux is good at the low-end server and desktop role, ans Jon and DH Brown state. However, I don't think your company is willing to run it's General Ledger, Web Server, or other critical systems on one of the Sys Admin's PCs. If they are, you need better Line-Of-Business people to whack upper management around a bit.
Linux is good. It's just not good ENOUGH yet.
This space for rent. Call 1-800-STEAK4U
RISCy Business sez: "Linux has no place ... as your 1,000,000 hit/hour webserver."
And just how many webservers get a million hits per hour?
Phil's post is exactly true - the hormonally addled 14 year olds who got posted on Mindcraft's "Linux Rage" page and made the entire community look juvenile in front of the world should read Maddog's article. At least twice :-)
The one thing about this post which redeems itself as being not a flame is the following line:
/* That's the way it is. Don't like that? Go work towards changing it. Change is good. But till things change, what I say will hold true. */
So this isn't FUD because it's open to the idea that Linux will improve over time.
The fact is Linux is, for lower end machines, excellent. Linux on an Intel against Solaris on a Sparc of comparable speed, for lower end stations wins hands down in ease of administration and (with the porting of Oracle, etc.) gaining rapidly in number of applications.
But until Linux develops a journaling file system (a real one: think AIX, not NT) or more scalability in SMP, clustering and HA (think VMS, not Janus), it will not take over the datacenter, which is where the real money and durability is.
That having been said, I disagree with the poster as to the "inevitablity" of it happening -- I think Linux, *BSD, etc will improve because you can't stop the desire to make free software better. Plus Free Software never will go out of business, by definition.
Give Linux 2-5 years to develop good HA and clustering. Give it 5-7 years more to get a reputation to compete with AIX, UNICOS, and OpenVMS in the datacenter.
Three Step Plan:
1. Take over the world.
2. Get a lot of cookies.
3. Eat the cookies.
I think you'll someday find that Microsoft is feeding the press with data to show Linux's weaknesses. That is how they 'compete'. Unfortunate for them, they aren't 'competing' with IBM, its Linux. Weaknesses will be patched quickly and tested by the community too quickly for Microsoft. They won't be able to wait for a liquid cooled CPU to become the norm so NT v5/2000 can beat Linux in future tests.
Fair benchmarking and reporting is not the norm in this industry. Recently IBMs Warp Server for e-business was hammered on InfoWorld (the link is now broken to the article....). It turns out the guy who wrote the column is the Senior Contributing Editor and Columnist of Windows NT Systems magazine. See "Hatchet Job"
Complaining/exposing a injustice is how we open the eyes of the unknowing. Example: At my stock investment club meeting last night, one member insisted that NT was faster then Linux in all cases and that Linux had a weak GUI. I booted OpenLinux v2.2 on my P120 laptop an he started questioning his beliefs. The rest of the goup was surprised at the polish they saw. All but one member are professionals in the technology industry though mostly embedded/realtime systems.
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
Yeah, when it doesn't hang your system. Mandrake is a good example of this.
Mandrake 6.0 has a hdparm line in /etc/rc.d/rc.sysinit, which optimizes HD performance on most systems. However, on some systems, it causes the HD to hang.
By booting without starting up init (linux -b at the LILO prompt), I was able to find and comment out the hdparm line. However, since the system hung during the first boot after install, the RPM database got trashed. I ended up having to completely re-install it again. That was a pain.
Mandrake should not have put in an unnecessary optimization in as a default. Looking at the discussion about this on their mailing list, they didn't seem to care, though.
That is true, but NT has a big head start. No, 8 machine failover is not stable, but NT at least ALMOST has a real presence in the high-end market. Linux hasn't even truly begun to concentrate on the "big iron".
If we are "going to get there first", we^H^H the Linux develop team have their work cut out for them.
----- if ($anyone_cares) {print "Just Another Perl Newbie"}
warn "Just Another Perl User" if $anyone_cares;
This is kinda ridiculous. Why is it that corporations are expecting the free(Linux) to be as "enterprise server ready" as the expensive(NT)?
Linux, as far as I know, wasn't designed with that in mind. It was originally Linus' hobby, for crying out loud. Also, it is a very young OS. It is right now good at what its intended use is... squeezing good use out of lower-end, and sometimes antiquated PC's.
I mean, try to make a web/mail server for your corporation with Windows NT on a leftover 486/16MB. Can't do it. But I did that with Linux, and it runs fairly fast for its hardware handicap.
But it has not yet been coded to do what NT is aimed at... taking over the high end server market. Linux developers (whom are gentlemen that deserve our utmost respect) are now beginning to seriously address Linux' lack in this area. I imagine that since Linux is already king in low-end servers, now it will turn its attention towards the higher-end market. Thus mindcraft and company have helped shape the development of Linux in a good way... instead of us madly shouting "unfair benchmarks!", we need to simply begin working towards what NT has already claimed... higher speed on the big iron. Hopefully we can do so without the bloat and instability that has plagued NT, though.
----- if ($anyone_cares) {print "Just Another Perl Newbie"}
warn "Just Another Perl User" if $anyone_cares;
If /. DID run on FreeBSD, that'd be a good reason to steer clear of it. /. is one of the slowest sites on the Net, period. The performance is absolutely laughable.
I think that most of Slashdot's performance issues are related to lack of bandwidth. If Slashdot is maxing out its pipes, no amount of hardware or software thrown at it is going to fix that problem.
Processing wise, Slashdot does a lot of dynamic content generation and is written primarily in Perl, which is interpretive. If memory serves, Slashdot runs primarily on a single machine and not a huge or expensive one, with a second machine that only serves up the graphics. Additionally, I believe that the database engine is running on the same machine as the web server. Serving up the graphics seems to be the slowest part of the whole system. If Slashdot used multiple graphics servers (preferably on seperate pipes), it might help matters.
I believe a significant boost in dynamic content generation performance could be had on Slashdot if it used a seperate database server box connected to the web server on a dedicated 100Mbit Ethernet. Moving to this sort of architecture could also allow clustering of the front end web servers. However, as I said before, if the bottleneck is bandwidth, that won't make much difference.
All in all, I think Slashdot does pretty well considering the limited resources it has had to work with. Now that it is owned by Andover, perhaps they will be willing to put some investment into infrastructure.
we need to simply begin working towards what NT has already claimed... higher speed on the big iron. Hopefully we can do so without the bloat and instability that has plagued NT, though.
How you can say those two statements together is beyond me. Do "enterprise" and "instability" go together? Face it. NT is not there either. We've got to make sure we're going to get there first.
Tell me where exactly would you go to get a computer with AIX if it's so great where can I get it in the intermountain west California, Idaho, Montana, Utah, Arizona, etc? How much does it cost? If it's that good and dosn't cost me my first born I might try it.
The death of one man is a tragedy; the death of a million is a statistic --Joseph Stalin
It could be a tactically designed method to lull people into a belief that they really are running apache and then just come out and say that they faked it and they were just getting stuff from IIS and disguising it.
The death of one man is a tragedy; the death of a million is a statistic --Joseph Stalin
John Haggerty
534 N Oakley St.
Salt Lake City, Utah 84116
The death of one man is a tragedy; the death of a million is a statistic --Joseph Stalin
Interestingly why are you using nt for printing with vmware? Is it an office application like word?
The death of one man is a tragedy; the death of a million is a statistic --Joseph Stalin
You know, I liked this article. It echoed what I've been saying for ages.
Linux isn't for mission critical large-scale. NT *DEFINITELY* isn't. If you want that, you'd better call your IBM RS/6000 VAR today, because sales are going to jump now.
Linux is not the be all and end all of unix. Period. It never ever will be, as it will more than likely collapse upon itself before we see ext3.
I use Linux. I've used Linux for about 4 or 5 years. I think Linux is great.
But it's not an enterprise OS. Period. Flat out. Never. It's good for small to midrange stuff, sure. Hell, our primary DNS server is Linux, as is our webserver currently. (It *will* be moved to an RS/6000 H70 as traffic increases.) Our secondary DNS server will be Linux. But our network monitoring system will be an RS/6000 43P-140 (aka Model 140) workstation running AIX. Why? Because if hardware starts failing in Linux, I'm screwed. AIX will scream bloody murder before it gets anywhere *near* the point of no return.
Linux has no place in the enterprise as an ERP server. It has no place as your 1,000,000 hit/hour webserver interfacing with SQL and doing dynamic pages. Period. Those of you who find this offensive, kindly contact a proctologist so that you may have your head removed from where it is. That's the way it is. Don't like that? Go work towards changing it. Change is good. But till things change, what I say will hold true.
-RISCy Business | Rabid System Administrator and BOFH
your company here.
shelby != ford
I just read their page at http://www.microsoft.co m/NTServer/web/news/msnw/Hotmail.asp and found that there was almost no information of substance! They didn't stint on the marketing information, though.
Linux is approximately the same age as NT, 7 years. Give or take a couple of opinions.
For the Hyperactive types among us:
Linux was not created to compete with anything. It was not created to 'overthrow' Microsoft, or anything so egotistical. It was created by people who needed it to do 'stuff', and do it reliably and cheaply and well. It still is, as far as I can tell.
Why all this talk about competition, Linux must beat NT, Linux is better, Linux this, Linux that? Why not just use it if it can do something for you, and not worry about what the rest of the world thinks?
Linux doesn't need market share to survive, folks, and it doesn't need acceptance by the enterprise-level corporations. Its Cast of Thousands who maintain various aspects of it will continue to do so whether Linux can beat NT on every benchmark or not... they will keep Linux going primarily because THEY use it, not because YOU use it.
Frankly, I think the gun was jumped in this 'race' with NT. Competition was created where none is necessary, or expected, or wanted.
Are you trying to imply the maddog is a lier? :p
But seriously, probably the best people to conduct benchmarks are not people actively involved in one OS or another. Considering the fact that maddog is the executive director of Linux International and the senior leader of the UNIX Software Group at Compaq Corp his opinion, although I would trust, might be considered biased by other groups.
Disclamer - Opinion of Person
I, for one would like to see a set of benchmark results for Linux that would help a person to make decisions about the configuration of hardware platforms for systems. I would like to see a set of test results for different types of system activity, such as compiling code, raytracing/graphics/visualization, file system access, network bandwidth, combined network/filesystem access etc. This set of measurements could then be run on a variety of hardware types providing the basis of cost/performance decisions in the implementation of systems. One could answer questions like how does a K6/233 compare to a P5/233 ? How does a P5/233 compare to a P6/450 ? How much difference is there between an ULTRA ATA disk system and SCSI? Does the difference change with different processor speeds ? Does 1MB cache make a difference with what I want to do ?
If implementing a cluster, does saving $2K per box make up for the difference ( providing money for more boxes ) if you use ATA and slightly slower CPUs, rather than higher end platforms ?
I would love to see a single ( and evolving ) location for this kind of info. Hell, I would love to work on compiling it. I have seen some sites with benchmark info, but nothing that seems to try to answer specific questions. The sites that I have seen present specific objective numbers but it was hard to derive any context for the differences between systems.
enough is too much
Actually Microsoft explains situation with Hotmail here
And homepages.msn.com is actually not Microsoft service but it's outsourced service ( don't remember who actually hosts it).
Actually it's run by TalkCity
yada yada ...
"plain evil"
yada yada ...
"division between rich/poor is widened"
yada yada ...
"free ride in this world"
yada yada ...
"dictatorial rule"
yada yada ...
"communists... Russia ... Soviet bloc ... democracy and freedom"
yada yada ...
"right to be angry"
yada yada ...
"stupid tests"
yada yada ...
"BSOD"
yada yada ...
"people who are paranoid"
yada yada ...
"industrial espionage"
yada yada ...
"arrogance and lack of tact about facts"
yada yada ...
"afraid to get hate mail?"
Is it any wonder that no-one listens to you? You are full of crap buddy boy! :)
When are you starting the terrorist bombing campaign, I bet you are an extremist right to lifer too.
So are you saying multiple network cards in a machine is "non-real-life"?
Maybe in your uni lab and garage fella, but computers are used in more places than that dude.
I think that you might be living in a "non-real-life scenario"!! :-)
ROFLOL!!!
Ever heard of hardware failure or application failure? The world doesn't revolve solely around OS stability you know... (And if you think your hardware is stable, try configuring it with a few GB of RAM...)
Call me a cynic (after all, I am paranoid for Linux's sake), but MS is probably holding off on moving hotmail.com and homepages.msn.com to Win2000/IIS until the optimal marketing moment sometime during Win2000 rollout.
"See, we've dropped Apache and are moving to our super-reliable, super-scalable Windows2000 with IIS. And we're now making that same power to your company for just $200!*"
--LP
"* Terms and conditions may vary according to the license purchased."
I liked the fact that instead of reading an
article about how messed up and skewed the
benchmark was, I read an article that suggested
things that can be done to better arm ourselves
the next time something like this pops up.
Good Job.
I wonder where all the hunters are today. --Daffy Duck
Who cares how Linux compares to NT? If I wanted NT, I would run it...I do sometimes, through VMWare. I use Linux because it's free, because it's open source, and because it's a good way for me to learn UNIX and programming in general. Linux is NOT going to be a Windows clone, nor is it going to achieve "world domination." I use Linux because I like hacking at it; I think it's great that Office suites and the like are beginning to be supported by it, but it's not a concern. If I wanted to use the best network server available, I'd use (and I do) FreeBSD...which Slashdot recently switched to, I believe (and which runs Yahoo!, M$'s own Hotmail, and pretty much every serious hardcore site you can think of). Linux IS NOT a windows killer; it is simply another (albeit WONDERFUL) alternative. If you think--and I hate Windoze bugginess too--that Linux is going to replace or defeat windows, you're F***ing nuts. Linux is a good, free way to learn UNIX and pragramming languages; now, with the support from HP etc, it is a syustem that people can actually get (Office-suite style) work done on. Nothing more. Am I concerned that NT beat it? NO. NT is generally crap--I run NT through VMWare for no other reason than it provides a convenient interface to /dev/lp0 for my school projects. If M$ came out with Office for linux, it'd be gone...but then they'd hopefully release Quicktime and MS media player for linux too :-)
here
People like Jon maddog Hall should be in charge of benchmarking. Then the world would be a better place.
Linux needs to become a good enterprise server. Firstly to run the enterprise applications (e.g Oracle) that are being ported to Linux.
Secondly to run them better than NT and beat NT in enterprise computing benchmarks.
The problem is that in order to achieve this, the linux gurus need to have access to enterprise hardware, which is *very* expensive. How many hackers can afford $10,000 for a development machine, let alone $100,000?
There are two potential solutions to this. One is serious investment by successful linux distributors (e.g RedHat & SuSe), the other is support from the hardware manufacturers.
This support could either be in the form of donations / loans of hardware, or by employing gurus to tune Linux for enterprise hardware.
If this does not happen, then M$ will keep tuning NT for high-end hardware and we'll continue to see results like the Mindcraft ones.
Yellow tigers crouched in jungles in her dark eyes.
She's just dressing, goodbye windows, tired starlings.
Window NT vs Linux sponsered by Microsoft...was anybody really surpirsed that NT won by such a large margin? You can't trust results like that when the sponsor happens to be the rival company.
Another test was done recently to conifrm the results from the Mindcraft test...and although NT still won, the margin was significantly less. THe details of why Linux lucked out are a bit technical, however, the cause of the problems were immediately identified. And there is even a Beta patch for the problems available on the site; it came out a couple of days after the problem was identified. In addition the problems will be resolved in the next Kernal release.
Just try and get that type of response out of M$, it takes them months to recognize that the problem exists, let alone come out with a fix...
"Imagination is the only weapon in the war against reality." -Jules de Gautier