Slashdot Mirror
Are Linux Transactions Slower Than Win2k's?
FullClip asks: "In the July issue of PC Magazine, Red Hat Professional is compared to Windows NT/2000 on basis of ServerBench, which tests the maximum Transactions Per Second (TPS) for a given number of clients. Red Hat 6.1 (when tweaked) matched the performance of Windows, but showed a terrible decrease in performance at about 24 clients to a weeping 20 % of the level that Windows was able to maintain. Somehow this disturbs me.
Doesn't Linux perform better than that bad in client-server environments? If someone can point me to an non-FUD benchmark site, it would be appreciated..." Is this yet another case where benchmarks have been skewed severely to show a deficiency that doesn't exist? Or is this another area where Linux needs improvement? [Updated 6 July 2000 2:15 GMT by timothy] You may want to compare this with the far different results reported by SpecWeb.
You people should never pull your Cox out at the last minute.
After a quick look through the benchmark (their executables aren't stripped), fascinatingly, it appears to make (extensive??) use of AT&T shared memory constructs (ie shmget, msgsnd, etc.).
I've never looked, but (given the age of the manpages) I wouldn't be surprised if this code isn't heavily optimized. I suspect this because, in my experience, very few free software programmers actually use these constructs (portability??). As a result, there wouldn't be a huge push for fast sV IPC.
Furthermore, I wonder about scalability of IPCs. A long time ago I ported a program (making heavy use of IPC) from AIX to Pyramid's flavor of Unix (does anyone remember what it was called?). The kernel default message queue size only allowed 8 messages while AIX allowed 256. As you might expect, this created a minor scalability issue.
You know, seeing benchmark results on this site really makes me laugh. No matter what, it seems like people on this site think that Linux is DESTINED to win every benchmark in the known universe and when it dosen't, the Test must be flawed, skewed, or paid off in MS's Favor.
My Personal favoritre was the Mindcraft tests, simply because in the end Mindcraft PROVED that it was a Linux problem. And they are still blasted for that. Thankfully Red Had did something about it and are finally getting the benchmarks Linux shoud have got in the first place, but it would have never happened if they did what 99% of the Linux community did and blamed Mindcraft.
Personally, going into a lull and believing that a particular software is untouchable is a disaster waiting to happen, because eventually its going to be overtaken by a competitor unless something is done about it.
-Slashdot: News for Linux. Nothing Else Matters.
No it wasn't.
It was a dual-cpu capable machine (as all E60's are) but only contained one CPU hence it wan't a mutil-threaded stack issue.
Did you notice they have instructions on how to install the program under Linux that are available in Microsoft Word format?
Check this out.
It's a SELF EXTRACTING WINDOWS BINARY ARCHIVE.
Fortunately, it's a self extracting windows binary ZIP archive, so I used "unzip" to uncompress it.
But then I tried to open it...
Makes AbiWord crash.
mswordview says "this is an unsupported word 7 doc, sorry
this converter is solely for word8 at the moment
"
which I found weirdly funny.
I finally got it opened with StarOffice 5.2, but gee.. what a pain.
How about plain text, or HTML guys!
Also, I couldn't get to thier license agreement link.
This is the saddest case of completely missing the joke I've ever seen on /.
It's just interesting that the SpecWEB tests showed Linux to be way ahead, yet ZDNet's tests (I still think ZDNet is in M$'s pocket) show Linux to be behind. Who do you trust? Hmm?
Sam: "That was needlessly cryptic."
Max: "I'd be peeing my pants if I wore any!"
>Basically, Linux people want Linux to be able to >do everything that Windows can.
:()
so far, so good. But then:
>They want it to be a robust server operating
>system. They want it to be an easy-to-use client >operating system. They want it to run everything.
So which one of these is Windows supposed to be able to do?
Guys, this *ISN'T* a conspiracy. As configured, the 6.1 box reached some sort of bottleneck. THIS ISN'T FUD for crying out loud. Sure, tests don;t mean much, but it failed it.
Do you guys/gals know how *BAD* we as a cummunity look when every time we're beaten on a test, we cry FOUL??
There will end up being a reason as to why this happened, but for crying out loud guys..
-- I'm the root of all that's evil, but you can call me cookie..
NT and security are two things that don't match. http://www.securityfocus.com/vdb/stats. html
War is one of the most horrible things a human can be exposed to. And one of the worlds largest industries.
WTF is the idea of using RH 6.1? 6.2's been out for how long? And the newer kernels from the 'updates' tree, like 2.2.16?
Sheesh.
Or could this possibly be a downside of Redhat... What? You mean Linux can sometimes have downsides too? Hmm, who would have thought...
I've never argued that Linux was perfect, but the nice thing about it is that when there's a deficiency in Linux, it can get fixed rather quickly. I guess a problem could get fixed rather quickly in Windows, too, but one would have to wait for Microsoft to A) admit the problem was there and B) put out a free fix for it. With Linux, odds are that someone will come out with a fix very quickly.
If memory serves, Linux had a patch for teardrop in 48 hours or something like that, and it took Microsoft a couple of months. That seems awfully long on Microsoft's end, but that's what I heard.
Yes, but the basic OS is still only single user, unless you're prepared to spend bucketloads extra on Terminal Server or Citrix stuff.
"The invisible and the non-existent look very much alike." -- Delos B. McKown
PC Mag's guilty of using an OLD version of an OS (Hey, they've shown themselves to be that clueless before in the past- why not continue the trend? :-)
Anyhow, the SPECweb figures are due to a machine that appears to have been running the pre 2.4 kernel, the Red Hat RawHide, and a nifty little high-performance web server called TUX that they GPLed that seems to outperform most of the stuff out there. Tidbit about TUX- it's a kernel extention not unlike the KNFS server.
I am not merely a "consumer" or a "taxpayer". I am a Citizen of the State of Texas
Just to clear things up, you're probably referring to Windows 3.1 when thinking about the 'single user, non-multitasking OS'. ENTIRELY DIFFERENT code base here; NT was built from the ground up (at least the important parts) separately from the Windows 3.1/95/98/ME family.
--
Never hit your grandmother with a shovel, for it leaves a bad impression on her mind...
Ahhhhh!!!!!!!!
Everyone is looking at this the wrong way. Why is everyone so concerned about speed? Are you aware that speed is not the only issue on computers? Gosh, the fact that Microsoft may have done something right somewhere is not that hard to believe. Why does everyone look at speed so much? Things like this make people forget why they chose Linux in the first place. That reason is freedom. That's the advantage over other systems. Having this freedom _usually_ produces better code, but it doesn't always. So what? With our freedom, we don't have to worry if RedHat puts out a better/worse system than Microsoft. If RedHat/Linux sucks, we can hire someone to modify it so it doesn't suck so much. And, if those modifications are worthwhile enough, we can sell them ourselves. The reasons most people choose Linux are freedom, openness, freedom from being vendor-bound, and customizability. Because of these things, we usually have scalability, performance, and security as well. But those are not the important issues. The fact is, you have the freedom to do things with Linux that you could _never_ dream of doing with Windows simply for legal reasons. This benchmark may be skewed. It may not. It may even be skewed in favor of Linux. But the fact is, I have my freedom, they do not. And my freedom is not something I deal lightly with.
Engineering and the Ultimate
> Why would you link to that? According to the very first table, Linux (aggr.) has 147 vulnerabilities over the four years listed, while NT has 146. Granted, that doesn't say much for Microsoft, but Linux is not a secure OS (most distros are as bad as Microsoft in enabling useless, potentially vulnerable services in the default install). Linux can be MADE to be secure, but then so can NT, so there's really no point to be made there.
Scan down a ways, you'll notice that for 2000, the top two are Windows 2000 and Windows NT. for 1999, the top 12 are all Windows (NT, 98, 95, IE, etc). Of particular note is that Windows doesn't show up at all in the 97 & 98 lists. Somehow I don't think it suddenly broke. I suspect it wasn't being reported on bugtraq. Just because it isn't reported doesn't mean it's not vulnerable.
On your second paragraph, I agree.
I wonder if it has anything to do with it running DB2, all of the other systems that were benchmarked were running Oracle 8.x. And the performance is soo vastly different for a little quad processor intel machine, I find it hard that it spec'd higher than any UN*X, even an Alpha Cluster, Sun's Starfire, HP's V class, and IBM's own RS/6000. If nothing else the sheer redundancy of the other setups would have me choose it over the intel setup on w2k.
And I'd also trust Bea Systems' Tuxedo for a TPM than I would M$'s COM+. Let's just say I'd rather use something battle tested when my job is on the line. (Yes I do use Linux at work, just not in my production critical back end database)
"If you insist on using Windoze you're on your own."
businesses aren't into the upgrade every week (or even every 4 months) paradyne
So what do they make of better-than-quarterly "Service Packs", some of which break their systems and lack of which leave "rape me" signs up on each network interface?
Betcha the results are miles apart if done with, say, Mandrake 7.1 - and another quantum leap if you plug a 2.4.0pre kernel in. In short, by the end of the year, nothing Microsoft does will bring those tail-lights any closer.
Got time? Spend some of it coding or testing
Here is another result posted on zdnet
The gentle swishing noise of reality vanishing out the door.
throwing up [...] is cake on IIS
Amazing what a bit of editing can do. (-: Sorry... back to the plot...
throwing up a simple vb or c++ COM object for scalability is cake on IIS
Whereas on Zope or PHP, the better design makes such kludges pointless?
Got time? Spend some of it coding or testing
>PS Did I tell you that Microsoft will release an
>operating system in the next 6 months that has no
>bugs at all, is as fast as hell, and has a 500kB
>footprint! So much for your Linux!
Nonsense! That Microsoft Linux version was an
april fool joke!
You obviously don't use or haven't seen a Caldera eDesktop system boot. It boots like the HP workstations boot, with a line of text describing the process and a status:
"[ ok ]", "[ wait ]", or "[ fail ]"
all wrapped up in a pretty GUI. HP did it in text but Caldera's would NOT scare any MS Windows user.
The login screen is pretty straight forward though many ludites.... I mean Window-ites wouldn't understand what they were logining in for.
IMHO
Funny though that at my July 4th party, 2 friends who recently bought computers said they paid $1500 for email and web browsing. They told me to shut up about the $99 IOpener I had showing photos of past events.....
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
Yeah, it's supposed to be a POSIX compliant system. Maybe we should call it X/Open Linux or something ? (-;
Remember, without a great visionary such as RMS as our leader, there wouldn't be any Free (as in speech) Software and the world would be a much worse place.
We'd have BSD and its associated license and variants ( like the artistic license ) with or without the "free" software foundation.
I would be very suprised if either OS had the lead in all aspects of webserving. If they're running on the same hardware, they're likely going to have similar performance - both systems have been developed by a lot of intelligent people, and are sufficiently advanced now that it's unlikely that they're just throwing performance away left, right and centre on stupid stuff. The areas that they perform well will probably be different due to design decisions they've made, but overall I'd be suprised to see a huge discrepancy between the two. Which one you want will always depend on what you want to do with them.
Context switches between processes take longer than switching between threads, of course, but the difference is far less under *nixes than under Windows. Interestingly enough, Win2K has much better process-switching optimization than previous versions and MS is now talking about (may have even released - I don't follow them that closely) a version of IIS that runs multiprocess instead of multithreaded to improve stability.
But yes, you're right - none of this does anything about the TCP/IP stack.
It's also worth noting that until June 30th, Microsoft Windows 2000 running SQL Server 2000 on a Compaq machine had taken the price/performance lead AND the sheer performance lead. Alas, I haven't read SPEC's rules yet, but their disqualified numbers are here.
Yes, I know... I found this out on the Register.
But it looks like IBM's toasted those disqualified numbers anyhow... cool!
Well, that explains the relative levels of stability and security of the two very nicely. 20 years (actually more) of proven technology == good, reinventing the wheel based on DOS == bad. Thanks for summing up :)
Your right to not believe: Americans United for Separation of Church and
In general the Linux model is to make forking a new process extremely cheap, so that you can just use new processes rather than threads. Threading hasn't been there since the beginning because there hasn't been as much of a need for it.
Your right to not believe: Americans United for Separation of Church and
The point is not "stability above all else", the point is that in OS design as in other things, you can either learn from the mistakes of the past or repeat them yourself. There's no reason to sacrifice stability just for a snazzy user interface, and there's no reason to sacrifice security just to get singing and dancing attachments in the mail. That's why 20+ years of design and industrial usage of an OS (if you consider all of Unix as one OS, which is a vast simplification) is a good thing, not a bad thing.
Your right to not believe: Americans United for Separation of Church and
http://www.zdnet .com/eweek/stories/general/0,11011,2290989,00.html
It basically shows that a C++ built com+ object running on Win2K smoked away every other platform (hard to believe I know..) Doesn't make a whole lot of sense but this seems to be coming from a good source.
I think MS has been doing a good job putting up some competition in the web server market. Apache is nice, but it's so simple to get ASP up and running on IIS.. (granted, PHP on Apache isn't bad either but throwing up a simple vb or c++ COM object for scalability is cake on IIS).
Many questions pop up for this test...
How much of the test relied on static pages how much of the test was dynamic pages?
There is a Linux kernel module to speed downloading static pages. It's just a remote file copy and there ends up being additional overhead when a user space application (Web server) dose the work. The same module passes dynamic pages to the server.
Linux however dose a decent job on dynamic pages.
If your website is graphics heavy then your proformence for Linux should be less than NT but if you have a lot of CGI,PHP etc type content you should preform better than NT.
The number of users (at once)... Can Linux or NT be trusted to such a high load environment? I'd think if you get 20 or more hits at a given moment you'd want to consider Solarus or other high end system.
To make it clear... It should take about 2 seconds (at worst) to get a page out. 20 people at once. So say half that for a normal load.
10 for every 2 seconds. 5 per second.
5 * 60 = 300
300 pages a minute.
18,000 pages an hour
18,000 for 8 hours a day = 144,000
[8 hours instead of 24 becouse people sleep]
Thats a lot of traffic...
My math may be a bit off but even with a few mistakes and bad assumptions you are dealing with heavy load when you expect a system to serve off 20 pages at a given moment.
if your dealing with an odd burst the system should be able to handle it given no new traffic.
I don't actually exist.
If you're going to base your decision on server software on how many millions of transactions it can make per second then perhaps you should try management.
Every E series Netserver I've used (45's and 50's) drops to a crawl once > 500Mb RAM is added. Up to 500, it's snappy, but the BIOS, and presumably the caching, or lack thereof, give the infamous Large Memory condition of Linux. I have not yet tried hurling a Gig at my LH4 beastie...
So if it's not positive about Linux it absoluteley *MUST* be FUD?
This ask slashdot is really clutching at straws... someone, please tell me it isn't true, tell me this bench mark is a bunch of crap?
I wholeheartedly agree. I applied for a job with a certain Linux-oriented company (which will remain nameless) awhile back. A job for which I felt I was seriously capable of excelling at. Sent in my resume and cover letter after some slight modifications in Star Office. What was my reply? "We have no need of people who use Microsoft software. We are an open source publisher". Where did the email come from? Netscape Communicator. I never really saw Netscape as a great bastion of open source despite the Mozilla project. Nevermind the fact that I was not given a fair shot at the job, this person reflected everythig that makes me embarrassed of the Linux and Open Source communities.
;-). See what the uptime and load averages look like. See how the OS's handle upgrades, patches, hardware swap-outs, etc. To make things interesting, don't put them behind a firewall. See if any of them get busted into.
Zealotry never accomplished anything good. If NT beats Linux, we pick apart every little detail of the test and find flaw everywhere. If the results are reversed, we cheer and sing the praises of happy penguins.
Open source is probably the best thing to hit software development. I believe it produces a much higher quality product. Also, let's have a look at a little OS history here. How long has NT been around? Unless I am mistaken, NT has taken longer to be 'competitive' with an OS as relatively young as Linux. For me, that speaks volumes as to the benefits of open source.
I would love to see NT take on Linux (or better yet, one of the BSD's) in a truely impartial, real world test. Personally, for a server, I would prefer FreeBSD. On that note, why have I not seen any big benchmarks tests comparing NT to it? Why is it always Linux? I could have sworn the last numbers I saw showed BSD on more servers than Linux. Perhaps we need a Road & Track-style head to head comparrison of the top 5 server OS's.
I would set up each machine as comparably as possible (on identical machines where possible) and advertise on the tech sites with the highest traffic numbers to hit the site as hard a possible. Hell, to keep all the OS communities content, let each vendor spec out their own server. I understand the need for consistency here, but consider that the vendors are picking what their product will run best on. Basically forcing everyone to put their product where their mouth is. Have some static pages for a day or two, add some graphics intensive pages, then to dynamic pages, and on the last week, mix it up. Anything goes, write scripts to hit the site, ping it, anything you can think of to beat the server ragged. After that round, move to the next OS. After this is done, hell, set up the 5 servers as ftp mirrors for something like Metalab for a year, something with tons of traffic (imagine an NT server mirroring Slashdot as part of the tests
I agree Windows has its place. Personally, I would hate to see Linux on everyone's desktop. Can you imagine how dumbed down it would have to become for that to happen? It's frightening.
/Sig/
If you're going to crow about security, at least back a secure OS, like OpenBSD, and in the meantime, pressure Red Hat and the other major distros to think more about security in their default installs.
I don't know of anything magic about 20, but everyone knows, the answer is 42.
Folks,
I'd hate to break it to you, but sometimes, one OS is better for one thing than another. Sometimes it's Linux, sometimes it's Windows (gasp), and sometimes it might even be something like CICS or MVS on a mainframe.
The ONLY way you can get a REAL benchmark, that absolutely applies in the case that YOU are interested in, is to do your own.
You might end up surprised.
I believe I've said this before, but about 1.5 years ago, we were making a major database/Server OS decision at work. A large enough decision that we were willing to put the time and money into some serious testing (It cost us about 500k to do the testing, partly because it took months). We expected to see one set of results, and came up with the opposite (BTW The opposite saved us money). The thing is, I have friends who have do somewhat similar tests (NOT as extensive) and they go different results than we did (They got what we expected to see). It all depends on what YOU are doing.
There are tradeoffs in every design. Heck, you know that when you build a database, an index speeds searches, but slows down inserts/updates, right? Well, the same thing goes with products. Sometimes you'll tweak one thing over another.
You can't really be an OS Bigot. I run Linux, NT2k, NT4 and Win98, and each has their place. Pick what works best for the job at hand, and work with it. When you get out into the "real world"(tm), you find one thing, you clients/bosses really only care about one thing, getting the job done, preferably for the lowest perceived cost to them. By cost I mean the REAL "total cost of ownership". The thing is, every company/person puts different values on various things, which can change the "total cost". If you client is a Linux shop, and has only Linux people on board, the cost of Linux is going to be less than say, AIX, or Windows, because they don't have to train/hire someone. If they are a windows shop, all other factors being the same, windows will be cheaper. That is why it's so hard to get a new OS/Program/whatever into a client. The biggest cost is PEOPLE.
-- 73 de KG2V For the Children - RKBA! "You are what you do when it counts" - the Masso
AC,
I think you and I agree! I too wish for more cross platform stuff, and if anything good comes of the Jackson decision, this will be it.
Right now, IMHO (and this is NOT a troll) Linux comes THIS close to being REALLY really for "Prime Time". It's STILL to hard to get configured on the average desktop (It HAS gotten a lot better, but...), and the driver support etc isn't there. It's polish issues, ease of software installs etc. Most of the stuff out there assumes you understand Linux. The end users don't WANT to understand Unix (or windows for that matter) - They just want to get their job done, and in general, that job is NOT using computers.
-- 73 de KG2V For the Children - RKBA! "You are what you do when it counts" - the Masso
...if NT beats Linux in a given area. I could care less. My Linux box as a workstation and my company's Linux box as a server OVERALL beats the pants off of our NT server. (Okay, so our 1GB RAM, 27GG SCSI RAID 5 NT4 SQL Server is faster than our 96MB RAM, 10GB IDE Red Hat 6.1 web server/SQL server/proxy server/etc., but we never have to reboot the web server!) I could care less if NT can do more TPS than Linux as a server to Win32 clients on some obscure client/server benchmark.
All that this tells me is that in one area, NT/2000 is faster than Linux. No one here is going to do any where near that many TPS.
This test does question the scalability of Linux vs. NT/2000. It shows that NT/2000 scales well to multiple users with their application. And it shows that Linux may have a problem scaling--but proves nothing.
If the above discussion of the TCP/IP stack of NT/2000 vs. Linux holds truth, NT/2000 should and would have an advantage in networking scalability, and thus client/server scalability, on a MISD machine (and possibly on a SISD machine).
If you Like Win2K, use it. But if you like Linux, use it. If you like Win98, then you must be smoking something (i.e. hallucinatory)!
To my understanding, the setup is 2 NIC's and one CPU even though the motherboard is capable of 2 CPU's. They feed 24 clients on each NIC and gets a performance drop with 24 or more clients.
The sudden drop in performance when the second NIC kicks in looks just like a simple and embarassing configuration error.
It even looks like the second NIC isn't working properly. The output of ifconfig would be very interesting reading for me.
The kernel is to my understanding 2.2.12 which should be in pretty good shape even if there are better kernels around now.
//Pingo
--- Linux or FreeBSD, it's like blondes or brunettes. I like both. ---
I mean, why not do a true 6-way shootout and let the chips fall where they may ...
Will in Seattle
If the NT and Unix versions of ServerBench are very different, it's possible that the NT one is just better written.
One way to use ServerBench even if that's true would be to compare ServerBench scores only between different flavors of Unix.
I just checked the ZD site, and couldn't find any published ServerBench results for Unix. They don't have a central repository of results, it seems, but a search for "ServerBench" in all of zdnet showed very few hits.
You can find an unbiased summary of NT-versus-Linux benchmark results at
http://www.kegel.com/nt-linux-benchma rks.html
> .... point me to an non FUD benchmark site ....
IMHO there is probably no such thing as a non-FUD benchmark site.
Trust nobody elses benchmarks. Run your own tests and make up your own mind which is best for your purposes.
-- MartinG To mail me: echo kewyjlcxyzvjfxbqwh | tr bcefhjklqvwxyz
Come on please, WTF? So the CEO of MegaCorp reads Slashdot and was just about to switch millions of computers to linux, but now because he's so confised about what the latest version number is he's not going to. I think not. Plus, sorry to dissolution you but "careless talk costs lives"? I think that I can categorically say that no-one has died in the Operating System Wars. Get a life.
Why are you posting anonymously? Don't you think it's an abuse of the system. Or are you scared to take responsibility for your own comments?
I'm sorry to have upset you so much. OK, fair enough that you don't want to get an account. I suppose that's your business. I respect your wishes in this matter. It is just my opinion that when joining in a conversation it is manners to identify yourself. I am also of the opinion that when posting under an identity people might be more likely to think about what they are posting. Ever thought what the coward part of anonymous coward means? Just my thoughts.
Sorry mate, that (almost) is my real name. Phil Meadows from Reading, UK.
Sorry. I really wasn't trying to get into an argument. This isn't meant to be argumentative either, but out of curiosity why don't you want a slashdot account?
About eight years since early versions of Windows NT, I think.. and Microsoft spend billions upon billions of dollars on research, and have tens of thousands of top class technicians working hard on their products. Whereas Unix is still pretty much the same as it was 20 years ago, apart from a few cosmetics. It's good that Sun is making some progress with Solaris though.
This is what open source is about. Find the bugs in the Linux product (it runs too slowly with multiple CPUs under heavy load), so we can fix them.
We've seen from the SPEC benchmark and RH's TUX server what kind of results you can get when you concentrate on solving the problem.
When this kind of thing gets folded into the mainstream linux distros, then Microsoft will really have something to worry about (like it doesn't already)
So keep the benchmarks showing that Linux is too slow coming..
I gots ta ding a ding dang my dang a long ling long
I agree with several of your comments on windows vs. linux. I really hate seeing all this attention on linux coming from M$ turncoats. I for one did not jump on the "linux for everyone" bandwagon; if it fits the bill then use it.
;-)
.02
Linux' strength comes from the fact that there is source code available and that one can get the whole system (source code and all) for free. Persons can also contribute to the system and have other folks help them out finding bugs, adding features or whatever. The BSD's don't make this quite as easy. Solaris, AIX, SCO, Winblows, etc this isn't even an option. Those systems are targetted at people who don't want/can't provide their _own_ solutions. They get a contract (warranty, license whatever) and the vendor deals with their problems.
Unfortunately, the mass media have tapped into the "Linux craze" and fluff it more that most of us want. That's their nature; these are the same folks who don't know the difference between a hacker and a cracker
We can't really do anything to reverse all that has happened, and would we really want to? I think its great to see all these new linux-related companies. They've legitimized using Linux in the workplace; I just don't agree that it needs to replace M$ products at all levels. However, if it makes a better WINS, DNS, Web, whatever server then lets use it. I don't want Joe Dontknownothing buying a system installed with Linux, "just because." M$ have allowed people to be dumb for too long and I personally don't want to deal with those folks. I just want to deal with the people that can help me get _my_ job done quicker/easier with more flexibilty.
well, thats my
Bob
It only took you that long because you are an idiot.
let's have a purile 400 message discussion* while we don't find out!
Here's a better idea: You have one and I'll set my threshold higher.
Don't fight it son. Confess quickly! If you hold out too long you could jeopardise your credit rating.
Why is it so very hard to even think about a Windows system being faster than Linux in some area? I mean, really... I love Linux and it's been my main system for over six years now; it's an excellent (best?) desktop and small business UNIX. But there's areas where other systems, possibly including Win2K, are just better. Microsoft don't employ a whole lot of idiots and it's seems perfectly reasonably to me that they're capable of creating an OS that can give these kind of results... It's terribly arrogant and more than a little naive to think that Linux or any other OS will be brilliant at everything. That's just not how the world works.
9.5 times out of 10, the mainstream media talks out their ass. Why anyone in THIS crowd hasn't figured that out yet is beyond me. (You are all a pretty smart bunch, c'mon, its the media)
As long as everyone keeps pushing what Linux can do, and keeps adapting it to their needs, it will continue. Will it have 75% of the desktop market share? Probably not, but who the hell cares? As long as the O/S that we all love is still being put to use, and that we can continue to tinker with it, who the hell cares what Windows is doing?
Hell, just look at the Amiga crowd. Their O/S is in hibernation, yet they still make it do cool stuff. At least without having a central company in charge of Linux makes it a lot more likely that it will be here in 20 years.
D
I sig, therefore I am.
Ok... so this benchmark is skewered in favor of Windows.
The other recently posted benchmark showing that linux was so much faster as a webserver was skewered in favor of Linux.
I'll just dismiss both of them as meaningless now.
-- "So they told me that using the download page to download something was not something they anticipated." - Bill Gates
VOEST is one of Europe's top producers of high quality steel. They also build other steel factories all around the world, from Nicaragua to Lybia, from India to South Africa. Only in Linz (this is where they have their main factories) they have such a big network with 60 km of network cables and more than 4000 Wintel clients (+1000 terminals). This network formerly had a big Amdahl mainframe as server, and several Alpha 10000 or something. They're currently about to replace these servers with cheap Linux computers, not because of the price or TCO, but because of - stability - ease of administration Linux IS an option for large, really large mission-critical systems.
A monkey is doing the real work for me.
Well, they did say that they got Red Hat to give them tuning advice: if RH can't customise their distro correctly...
--
Cheers
Cheers
Jon
No, NT is not a microkernel operating system in the accepted sense of the word, as you yourself say later. Microsoft describe it as a macrokernel OS - its design is informed by the theory behind microkernel OSes but, as you go on to point out, it breaks a lot of the "pure" microkernel principles in order to be usable on (then) modern commodity hardware: when you're designing for a 33 MHz 386, doing frequent switches between ring 0 and ring 3 becomes expensive, and as has been pointed out many times, if you lose, say, your user-mode filesystem you're, realistically, just as fucked as if the entire OS crashes.
The original NT design featured a POSIX API layered over a slightly extended OS/2 API. The second implemented both the POSIX and OS/2 APIs directly in the NT executive (the bit you keep referring to as the microkernel). The third design, and the one that emerged from the labs, layered both OS/2 and POSIX as protected subsystems over the executive.
The original NT executive consisted of the underlying microkernel and I/O Manager (file systems, cache manager, device drivers and network drivers), sitting on top of the HAL and supporting the Object Manager, Process Manager, Security Reference Monitor, LPC Facility and VMM with the rest of the system services being dependent on these components. All these components ran in kernel mode. The remainder of the OS (Win32, POSIX, OS/2 and security subsystems) ran in user mode.
In NT4, the Window Manager, the GDI and the video drivers were all moved from CSRSS.exe (the Win32 subsystem) into Win32k.sys, a kernel-mode device driver. This, particularly the moving of the video drivers, is the main cause of increased instability in some NT installations: some companies (Matrox are, IMNSHO, a shining example) have consistently produced high-quality and stable video drivers, while others (who shall remain nameless - you know who you are) have repeatedly produced piles of shite. This said, moving the video subsystem into the kernel has become a convenient rod with which to beat Microsoft's back: NIC and SCSI drivers have run in kernel-land since the off and while I've seen some shocking drivers attached to no-name NICs, I've also found that most quality NICs (i.e. Intel and Adaptec) and SCSI adapters (i.e. Adaptec) haven't impacted on the stability of my NT boxen.
NT was designed for VERY high performance I/O.
And this is a bad thing in a server OS?
Win2K is definately not designed to be extendible and general.
Could you come up with some justification for this comment?
While WDM may add a lot of overhead to the driver interface, that is not NT's native driver model. Microsoft added WDM to allow drivers for Win98 to work on NT.
I keep coming across this from people who should know better. As of Win 98/NT5, WDM has become the standard driver model for Microsoft operating systems: older NT drivers may or may not run (specifically, most video drivers will run, but most others won't) but all new drivers should be written to the WDM. And no, this does not mean that the same drivers will work on both Win98 and Win2K - in fact this is highly unlikely, as a moment's thought should make clear.
For example, Windows 2K has DirectX class integrated into the HAL.
And where exactly should the direct-to-hardware API be. It has to be on the HAL, since only the HAL can talk to the metal. Get a grip!
NT also runs all services in kernel mode.
Absolute cobblers: drivers are run in kernel mode, services run in user mode. Check your facts before posting.
The NT microkernel globs up a lot of services that should be in the servers, which furthers performance, but makes the microkernel less general and less extendible.
And once more: NT isn't a microkernel, so spare us the lectures from Operating Systems 101 or whatever book it is you've read this week.
Sheesh - if there's one thing worse than zealots, it's uninformed zealots.
--
Cheers
Cheers
Jon
Bzzt - wrong. Under NT at least, threads are an essential aid to making server software (i.e. services, COM servers and, particularly, MTS/COM+ components) work at any sort of speed: calling CreateProcess() is incredibly expensive (unlike calling fork()), so the standard NT model is to start a single process that has one or more master threads and then either spawns worker threads as necessary or allocates worker threads out of a pool. As a measure of its efficiency, SQL Server (all versions since 6.0) running on commodity Intel hardware has always done more TPC-Cs/$ than anything else out there - generally Oracle on big Sun or IBM iron held the top TPC-C numbers but SQL Server owned price/performance.
The other big advantage of threads under NT is that you can, if you want (and you have some very interesting requirements), replace a multithreaded app with a fibrous one and do your own fibre scheduling which isn't, AFAIK, an option on Unices (certainly I've yet to find a Unix which allows one userspace process to schedule other userspace processes)
--
Cheers
Cheers
Jon
Has nothing to do with expense...
You simply _can't_ do it, because Microsoft will not allow you to and publish the results. Read your software license agreement or EULA for _any_ Microsoft OS or product. Oracle and many others use the same language.
Bottom line, we have to live with skewed, FUD-soaked 'benchmarks' that the vendors then use to backstop their lies, because the EULA, the DMCA, and UCITA won't _allow_ the playing field to be level, and won't _allow_ us to be "free to innovate." The very things that they have lobbied for to protect _their_ "freedom to innovate" stifles ours!
Ironically, it is prompting _true_ engineers and scientists to come up with alternatives (and making them rich as Croesus in the process) that will mark the death of these companies. They are beginning not to matter. Would _you_ go to work for Microsoft or Oracle as a programmer, at their present market valuations for 1/3 the going market rate and the promise of stock options in 12 - 18 months?
Remember guys, this is Amerika. Just because you have the most votes, doesn't mean you get to win.--Fox Mulder
More importantly, we _cannot_ do a 'benchmark' or performance test of any kind, because of the EULA, the UCITA, and the DMCA.
We must live with the results of tests like this, because the companies involved will not _allow_ us our freedom of speech. We gave that away when we loaded their software, whether we read the EULA or not!
Remember guys, this is Amerika. Just because you have the most votes, doesn't mean you get to win.--Fox Mulder
...and then you will get an e-mail (followed by a registered letter) from Microsoft's lawyers, telling you to take the site down, in accordance with the EULA...
If you do not comply within 48 hours (may be 24 now) your home and place of business will be raided by armed Federal Marshalls, Secret Service and FBI, accompanied by local law enforcement, warrants in hand. These teams will be backstopped by similar ones hitting all of your ISPs and their mirrors and proxies, demanding the site be wiped and that the team be given all back-up tapes and copies thereof of the site, as stated in their accompanying warrants.
Live in a foreign country? Only difference will be that your teams will include members of your foreign ministry, your commerce ministry, and probably your 'internal security' forces, and the warrants will be in the national language of your resident country. You will be arrested for contempt of (an American) court, violation of the UCITA (via the WTO) and various intellectual property and copyright/patent laws.
Wanna do it to Oracle, instead of Microsoft? Won't matter.
Go ahead...make their day...post the results of your benchmark in violation of criminal law.
...or forget about it and use an open-source OS and software that is GPL'd (and therefore you already own and are free to do with as you will).
Remember guys, this is Amerika. Just because you have the most votes, doesn't mean you get to win.--Fox Mulder
Nope, don't really want my website torn down, my home raided by the Secret Service, the Marshall's Service and the FBI and be faced with criminal (violation of the UCITA, and the DMCA) and civil suit (violations of the EULA I agreed to _before_I_bought_the_software_ (UCITA again!)and my existing NDA's (in accord with my attorneys' advice)). It's why I post a personal, rather than corprate webpage in my (open!) user profile, and why I post as myself rather than an AC (apt name!). It wouldn't matter, by the way, whether I was in the US or in a foreign country (except maybe the 'Bad Seven'), I would still be subject to prosecution/persecution.
.dll's and unmunge them, using Microsoft's or GNU tools (Cygnus). Some of the names of the .dll's (in \Windows and \Windows\System 32) even hint at the facts, depending on the product(s) you have installed. You may also gather facts by perusing freshmeat or appwatch for translation utilities and looking at the source. Finally, go do some corporate research on EDGAR. That's where I first saw the 'public' trail begin to manifest.
You, however, as I stated above, can find the
Also, even if I did post, would you believe?
BTW, is it raining there in Redmond?
Remember guys, this is Amerika. Just because you have the most votes, doesn't mean you get to win.--Fox Mulder
...and did you see the cost/tpm-c (not to mention the fact that no Linux or *BSD h'ware/s'ware or systems is represented).
Remember guys, this is Amerika. Just because you have the most votes, doesn't mean you get to win.--Fox Mulder
Nope...
... (technical people) that are tired of being used to sell product and certifications like Amway salespeople / customers," (from http://www.osopinion.com/Opinions/DougChick/DougCh ick6.html), the closed-source models of development and business will be exposed for the lies and trash they are, and we all will be better served and "free to innovate."
When the benchmark is closed-source (ie. binary, precompiled, no source provided) against a closed-source OS (ie. binary, precompiled, no _true_and_complete_ source provided) vs. an open-source (ie., souce provided, including an analysis of _what_ the benchmarks are testing for and _how_ they are attempting to do this) benchmark against an open-source (ie., complete source provided) OS (could be Linux, or any of the *BSD's, or any other _fully_ open-source OS's), the activity is _truly_ FUD on the part of the closed-source players.
Benchmarking, done accurately and fairly, is an exacting discipline of software engineering. When it is done only for marketing pablum purposes, however, it _is_ FUD. Face it, dude (since your user profile is closed, unlike mine, I have no choice but to use the generic label), you and I cannot 'benchmark' any of Microsoft's OS's or other products (including all of Back Office) because of the restrictive software license(s). More accurately, we cannot run any kind of tests (industry-standard, such as TPC, or otherwise) and then report the findings and results publicly. We cannot test these OS's or other products (Microsoft, Oracle, et.al., all closed-source), and then present the results _even_at_a_scientific_or_professional_conference_ (such as the ACM SIG meetings).
Ironically, the above _facts_ support the open-source model, (since there are no restrictions on the software science and engineering communities) and lead to the proliferation of open-source solutions to software engineering problems. It is this process that will eventually _kill_ the closed-source business model. When academic and professional software developers, purchasers, scientists, engineers, and users begin to realize that the closed-source models of development and certification turn them into nothing but "a bunch of
Remember guys, this is Amerika. Just because you have the most votes, doesn't mean you get to win.--Fox Mulder
If the pc world was Betel (or BePowerPC) rather than Wintel, we would all be bitching about evil Be Corp. Some people might find Be to be friendly and open(ish) now, but that's only because they have to act that way just to survive. Microsoft does not act the way they do because they are evil, that's just how you maximize profits for shareholders.
Shareholders are even worse engineers than managers. Software wants to be free, and so do I. For most of what I do, linux is great, and I can wait for a threaded IP stack, etc. However, I also admin HP-UX systems that fill 3 or 4 2meter racks, linux is not ready for that kind of hardware, but it will be soon. And when it is the world shall be debianized!
The current Slashdot moderation system is made by gay communists!
Instead of getting all wrapped around benchmarks (let alone closed-source benchmarks!) we would be better off paying attention to the efficiency of the server program:
(by Philip Greenspun, from So you want to run your own server, which is Chapter 8 of Philip and Alex's Guide to Web Publishing)There is always need for improvement and that is why I like linux as much as I do. It is not stagnant like some other OSes. As for the claim, nothing ever stays at the top, but I doubt that a single corp. out did a world community of developers.
I also see that not everything submitted is the same, so how can I be certain that this combination is the best. The best way to prove this without a doubt is to use the a hardware system with various OSes and Distributions with a DB. Keep in mind that TPC depends upon hardware, OS and the DB that is using. From what I say was half done. No real conclusion can be made. Sorry
. Threading wasn't invented in universities, and MIT didn't write the most-promogulated implementation; Microsoft did, and first!
Who said that?
Moreover, I'm glad you cleared up everyone's misconceptions about multitasking - I'll never look at pre-emption the same way again! Damn those researchers for lying to us, and damn those vendors for selling us multitasking systems since the mid 70's!
What are you talking about and who are you responding too?
I could care less what any of you use, I'm a BSD man myself, but your post was neither insightful, informative, interesting, or funny. It wasn't even on topic relevant to what the parent had said. Why don't you just admit that you took this opportunity to flame an MS user trying to get some cheap karma?
Rev.Marc
I do not understand what the benchmark is trying to measure.
They claim to be measuring transactions per second. The industry standard for measuring transactional processing power is TPM (per minute), and the T applies to ACID transactions, which have absolutely nothing to do with application serving. See the Transaction Processing Council for more info.
Mixing apples and oranges in an attempt to confuse a few hapless MIS manager types.
I for one never disagreed that Windows is a fine application server. If nothing else, there's lots to serve. But as far as transactional processing, I would not bet my MIS career on M$
The numbers are impressive, but available 12/07/00? I guess that makes it officially vaporware.
I imagine by then Sun and Oracle will have cooked up a fine response.
I guess M$ did this so they could be anywhere near the top in anything but cost/txn.
Actually I have, I just was thinking of most linux distributions.
It's an easy to use client>operating system. You can argue about all the bugs it has and it's limitations, but a lot of computer illiterate people use it fairly well, the same sort of people who would shutdown a Linux box if they saw it boot because they'd think it was doing something bad.
What should be looked at here is not whether the test was rigged, I seem to recall Mindcraft being vindicated a while back for that.... the real issue is that even if Linux does do as well, there's always room for improvement 8)
Instead of griping, just get to work and make it better than it was, that way nothing is harmed, but a lot is gained.
Name calling, hardly a way to prove your point. Opinions are fine, but state them as such. If you have a problem with the way an objective test is performed you should qualify your arguement with specifics.
As for both operating systems being designed for the same market, where did you get that? Even MS doesn't those kinds of claims. And last I checked you had different distributions of Linux, frequently customized to different markets (handheld vs server). I certainly hope your not going to make the claim that Linux is designed for mainstream use, even in office desktop environments. Last I checked, most *nix users took pride in the fact that their product was /not/ meant for mainstream use.
As for AIX and Solaris, why would those be brought up? They should have been tested that's why. Do you sincerely think that the only Sun web servers out there are the 6 figure high end very scalable variety? As for comparing Linux to Solaris, do you really think that MS is the only competition to Linux? You said it yourself, work is being done to make Linux more scalable. You can't say "next years version will squash everything". You have to compare to what exists now. Just because you did not see the results you wanted does not mean the test was invalid. Constructive criticism should be taken as just that. If something doesn't perform as expected, you learn and grow from it.
Now that is funny...
.02
Linux, MS, MS, Linux.. the battle rages on.
Well, As long as the battle rages on, my work will sit comfortably back and not use either. *sigh*. I want linux here, but as long as it is compared to Nt/2000 it wont be...
I need it to be compared to the real OS's of development. HPUX 11.0, Solaris 2.6 (and up)... And all thier nice real time friends. Stability, I/O, and all the rest are very very importal!!
So, keep comparing linux to Windows and forget the reason linux could win, if they re-merge all the other unix products out there. If HP, Solaris, and BSD got together with linux... no stopping it at ALL!!
just my
I can program myself out of a Hello World Contest!!
Hi, These test results are very clear. Also look at the FMStocks application (www.fmstocks.com). This application just blew Oracle/Netscape and all the other away on the Transactions/second benchmark. Micro$oft is doing something right this time. Originally I was more UNIX-LINUX oriented, but for the past five months I've worked with the Windows DNA platform (Windows 2000 Server, MSSQL Server 7.0, Visual Studio 6) and I must say that I am impressed. With very little effort you can create multi-tiered, highly scalable en reliable web applications. The development process is very efficient, because using MS Developer Studio you can debug all the way from the presentation layer (Active Server Pages) into the MSSQL Stored Procedures. Try to do that on a LINUX platform... The bottomside is that you'll have to use M$ software all the way, but fortunately using XML, SOAP and the forthcoming Host Integration server you can communicate with any system you want. I just want to say that many of you should open your eyes and give M$ a chance. Some of their departments do make fine stuff.
Aches = AIX; Pains = Panes = Windows NT
GNU/Linux gets rid of all your AIX and Panes.Will I retire or break 10K?
Shush! This is a linux newsgroup! Posts praising MS are seriously off-topic here!
;)
I haven't read the original article in any detail, so I'll state the assumptions on which I'm working up front:
1. Each subnet they used to generate requests had 24 client machines
2. Each client machine produced about a dozen (12) requests per second
3. They, at best, set Apache on the server to the highest default child process count (255 or thereabout)
4. They saw a "tremendous dropoff in linux performance" when they added the 24th client machine, continuing as they added more machines
Now, here's the analysis part, based on my experiences administering, hacking the internals of, and benchmarking Apache. Yes, I did this for a living at GeoYahooCities, and later as an independent contractor, so I do have real-world experience to confirm my suspicions.
First, note the default internal Apache limit on child processes. I don't know what version of Apache they used, but the default limit I remember was somewhere in the area of 256. Now for some math... 24 clent machines, about a dozen requests/second each.. 24x12 = 288. Once all the Apache child processes are busy, all requests get stuck in the accept() queues of the tcp/ip stack, waiting for one of the child processes to finish what it's doing and serve the new request. So, if they're seeing things slow down compared to IIS after this point, I'd say the culprit is the default child process limit. Note that this is configurable at compile-time, so it's not an inexorable problem. And it certainly looks like the web server is, in this case, the culprit, rather than Linux as such.
Disclaimer: No Apaches were bashed in the writing of this comment. Any similarity to Apache-bashing, whether flamebait or trolling, is purely coincidental.
One man's religion is another man's belly-laugh. - LL
perhaps they've removed it because the tests were done using a BETA version of SQLserver 2000. All software used in the tests currently on the list are released versions.
--
Never underestimate the relief of true separation of Religion and State.
That's a story in Douglas Hofstadter's book "Godel Escher Bach, An Eternal Golden Braid". The particular brand of computer and operating system they were talking about was not mentioned.
All I want is a secure system where it's easy to do anything I want. Is that too much to ask ~~ Randall Munroe
RHL 6.1 is old, but so is NT4 sp6 and it outperformed both Win2K and RHL 6.1. Doesn't that crap on the "it's slow cos it's old" argument?
All I want is a secure system where it's easy to do anything I want. Is that too much to ask ~~ Randall Munroe
Well this benchmark done by CNN in january shows the same problem.
S .idg/index.html
http://www.cnn.com/2000/TECH/computing/01/24/NO
That when linux is subjected to a heavy load it fumbles and the performance drops. The article is pretty in depth for cnn. I hate people talking how 2.4 will fix everything, 2.2 surely didn't. Thats like saying Windows 2005 will be better you'll see! Wait for the testing before opening your mouth, it hurts credibility.
Only the State obtains its revenue by coercion. - Murray Rothbard
I'm suprised you weren't bitchslapped for speaking the truth.
Only the State obtains its revenue by coercion. - Murray Rothbard
That Adaptec RAID card is not supported by linux.
Only the State obtains its revenue by coercion. - Murray Rothbard
It seems hard to me to believe that Windows, that wasn't even really a multiuser multitasking OS until not too long ago, is now smoking the hell out of Linux at what the Unix comminity has been doing for years...
Eh...
Perhaps. I dunno, I run both. I get better performance from Linux. Just speaking from my experience. I really don't HATE M$, as many people here do. Also, Unix was designed multi-user, and Linux is pretty much a Unix knockoff. What I'd really like to see in this benchmark is BSD, but since Linux is the flavor of the month *sigh*
Eh...
were you born a GNU/Linux God, or did you have to learn it?
I had to learn it.
To make ammends for your wrongdoings, I suggest you go onto alt.linux and answer some newbies' questions
I have done in the past. (Although more experienced Linux users usually give better advice than me) Thats because these are people who are willing to learn something new, and aren't scared off by the idea that the kernel version number might be different from the distribution version number.
Thanks. Nice to know someone appreciates my sense of humour.
your antics have probably made someone who was considering making the switch to GNU/Linux decide to stick with Windows, because "everything is so much simpler in Windows".
They deserve everything they get.
Yes, I found it quite intriguing that no one has mentioned NT4's good performance compared to Win2k. Feature creep continues...bloatware expands.
Forvalaka
(Proud 'Win-32-Bit Console Mode' Semware Editor User)
Or could this possibly be a downside of Redhat... What? You mean Linux can sometimes have downsides too? Hmm, who would have thought...
What does the server DO?
How much I/O is done per transaction?
How much CPU-bound processing is done per transaction?
Does the server use threads? Is there a pool of threads (how many) or is a thread created for each transaction?
What OS calls does the Linux port use? Which socket calls? Is a new connection created for every transaction? select() or poll()? blocking or nonblocking sockets? etc, etc...
Is the data file opened with O_SYNC to emulate transactional properties?
How were the binaries compiled? Which compiler?
Given that this site seems to be very windows centric, how do I know that the Linux port is done properly by someone who know what she is doing? Can I see the source please?
It would also be nice to see some profiling data from the benchmark. Where is the bottleneck?
CPU utilization
Kernel/user mode utilization per processor
network packets per second
disk traffic per second
Also note that their term "transaction" is probably not a database transaction with ACID properties (as the industry standard TPC benchmark mandates). An ACID transaction needs to store updated data permanently on disk (not file buffers) before the transaction commits. I don't think the test does this.
I am implementing a database manager on Linux and there are some areas where I think NT has better capabilities than Linux (disk I/O for ACID transactions), but that is another story, since this benchmark probably doesn't do ACID transactions.
per@nospam.mimer.se
www.mimer.com
)9TSS
Let's say we take and build up a pretty solid PC from known good stuff. From personal experience and assorted reviews I would recommend the following.
Asus motherboard
Intel Xeon processor
Adaptec Raid Card
5 IBM SCSI Hard Drives in the Array
512M DDR Ram
10/100 3Com Ethernet Card
The basic parts can vary somewhat as I'm sure other folks have other preferred hardware. The important thing here is that the platform be relatively stable and quick for any of the tested OS's.
Build 2 of these PC's to sit side by side. One to act as the web server, the other to act as a database server. There's no point in putting both on the same box, since that's usually not a common configuration whether you're talking NT or *nix.
Develop a simple dynamic page for pulling in 10 records from 10 database calls on each page load. Maybe toss in a couple of inserts and writes for good measure. Each page that's presented should also be required to load up 5 30k graphics to further increase the load.
Step 1, load up NT 4.0 with all the latest service packs and patches on both boxes. PC #1 gets IIS installed, where PC #2 gets SQL Server installed. It's probably fair to say that most folks who prefer NT to web host generally tend to stick with NT on the db side as well. The scripting language here should be ASP, as this would be a more common setup for most NT based sites.
Step 2, slam that set up with a heavy simulated load from a number of external PC's. How many page loads can ya get per second, when does it start to break down, and all that jazz. In addition, the testing should also push a heavy load on the server for a 24-hour period with random bursts and sags.
Step 3, wipe the two test systems clean and start over. Using the identical hardware install the next OS/Web/Script/DB combo and perform the exact same tests. The clients performing the calls should remain whatever they started with, so these cannot impact the test.
Combos that I'd like to see in various combinations are...
OS: Win NT 4.0 / Win2k
Web: IIS 4.0 / IIS 5.0
Script: ASP / Cold Fusion / Perl
Db: SQL Server / Oracle / DB2
OS: RedHat / Mandrake / FreeBSD
Web: Apache
Script: PHP / Perl / JSP
Db: Oracle / DB2 / PostgresSQL / MySQL
Real world dynamic page loading, database handling, TCP/IP stack management benchmarks that would actually be of some value. Lose all those gray areas involving different hardware specs, different network architecture, and questionable motives.
Probably too expensive, and I also have to wonder just how many folks would want to see the honest results from a test such as this.
The line must be drawn here. This far. No further.
Read'em and weep
Well, I would like to, but seeing as how the next 8 systems down the list are running UNIX, I can't seem to find the tears. ;)
Chris
-- Humans, because the hardware IS the software.
Well it's nice to see that someone else is open minded in the geek community! I'm an MCSE, I hate Micr$oft products, but in this industry that's sort of irrelevant. I know that Windows has it's place in the work world. Linux is a strong OS, but Windows has market support and therefore has a key foot hold that makes it better than Linux. I use Linux at home for my server because I'll only ever have to reboot it if I upgrade the kernel. NT uptime on the otherhand is a FAR cry from Linux. At work we are a Micro$oft solutions provider. These NT systems need to be rebooted and have troubleshooting run ALL day! Any business would be far better off running in a Linux environment, but let's be real here. Companies hire consulting firms (like mine) to purchase equipment and setup networked environments in their offices. Hardware has NO margin thanks to crazy competition in the industry and installation of a system doesn't take more than a few weeks (a few months tops). If firms installed Linux or Mac based systems they would be able to install the system and come back on occasion for security patches and other general and unavoidable maintenece. However, you install a Micro$oft based solution and you can do what I do do and provide four 8 hour days of support on site because the systems constantly need attendance for some reason or another. Now the company is reliant on our firm for support and we make more money. Microsoft is easy to use and corporate users have an easier time adjusting to an interface that looks like what they use at home. It's certainley not the best solution, but I'm open minded enough to understand why it's used. I started using Linux because I was open minded enough to use it and because I support competition and would like to see the OS grow to be adopted in the business environment. I just wish some of my geek friends would admit that Micro$oft has a function and not to slag me everytime I say that I'm an NT admin by trade. ------------ Those are my thoughts not yours!
WURD!!
There might be a real economic interest in comparing two systems on a specific task, and those tasks are different on different platform/distro/OS combinations. Obviously Linux nowadays is not equal to Linux, since those distros are ways too different, but knowing, that Red Hat does not do the job is greate! Now we can test SuSE, Debian etc. and later we may find that Free/Open/Net-BSD's are nice too. Well I think we should start benchmarking big time. Because MS is strong and good, since most buissines aplicationes run on Win9* its just fair that Win2000 does fine (And quite logicaly to me). But most DB's, routers DNS Systems do better jobs on certain Unix (or alike) systems. If we can proof such advantages with some numbers for the suits, we might just sell our idea just that much better...
42 cows on a 42km road on their way to 42.org
Benchmark tests are made to test exclusively one feature of one particular product under certain arbitrary conditions. The only way to really know is to try it yourself.
"If I were to ask you a hypothetical question, what would you like it to be about?"
Stability
We rebooted our NT Web Servers every night to achieve Stability (all 15 of them). That's first hand experience. By the way, they'd still crash during the day.
Speed
Define serious benchmark. If 'serious benchmark' means that it shows Windows as faster than Linux, then sure, Windows is faster.
Price
Linux administrators may cost more (I don't know this for sure, but I'll concede it to make the point), but in the aforementioned server farm, we had to staff the joint 24x7 for the nightly reboots. Think that was cheap?
Security
You must be kidding. Are you trolling?
I sure don't think Linux is a panacea, but I *know* Windows isn't.When the article shows Windows has an edge (however slight) then it's "FUD"... when it shows Linux beats the pants off Windows then it's the "truth".
That's right! You read the subject! I can see it now... I will get two machines: one running linux, and the other running a pre-release version of windows 2k. Before I begin the test, I will indiscriminately remove some system files from the win2k machine. Naturally, I won't tell anyone I did this. Then, when the windows machine crashes in the test, my headlines will read "Linux performs more than twice as fast as Windows 2000", as in infinitely, because Windows didn't even work!
Then everyone who claims that all the tests that say that Windows 2k outperforms Linux are flawed will have another benchmark to tell them exactly what they want to hear!
The point of my rant is that there is obviously no point in posting these benchmark results on Slashdot other than either starting a flame war or telling all the people who use Linux anyway exactly what they want to hear and believe.
Friends don't let friends use multiple inheritance.
check the newswires. ESR just suffered a heart attack at the Merrill Lynch technology convention. I've submitted the story, but it's on (a href=http://www.reuters.com>Reuters right now.
-- the most controversial site on the Web
I don't know where it comes from, but in my office our Red Hat 6.2 Application Server (serving Oracle applications) is humming right along at 46 users. We had a windows box for a little while, but it crashed too much, go figure.
"The words of the prophets are written on the Slashdot walls."
He can keep the karma, who needs it?
Yes, I was flaming. And why not? this is slashdot...
as for the content of the thread, I was just throwing noise around, around the subject of the original poster insinuating that NT was first on the scene with tasking/threading, quote "years and years", which annoyed me. The operating system problem is an old one, and although some advancement is happening in modern commercial OS'es, NT is hardly a shining example. But I disgress.
I was pointing at how old those two things are, so a single, closed-source vendor could hardly claim 'first' on either. Since I hardly have read access to the NT source tree, I can't tell you when they were added, so that's hardly a defensible position. But I should have jumped him on pthreads, since with them Linux has been threadable since 1.2...
Um.. yeah. And it shows, too, don't it? I mean, all that stability and all? Boy, all this time I thought NT was a munged knock-off of VMS, with implementation details following the latest breeze (remember when they said it was going to be Mach-based? that was a good one).
But it looks like I was wrong, and you are right. Threading wasn't invented in universities, and MIT didn't write the most-promogulated implementation; Microsoft did, and first!
Moreover, I'm glad you cleared up everyone's misconceptions about multitasking - I'll never look at pre-emption the same way again! Damn those researchers for lying to us, and damn those vendors for selling us multitasking systems since the mid 70's!
please! Enlighten us more!
...and you can take part in discussion
moreover, not once did I mention Linux, nor allude to it. Learn to read.
I couldn't care less what your meatspace name is or is not. This forum is an electronic one, and handles are perfectly acceptable. But an identity is, in fact required, not 'Anonymous Coward'. Register, and you can be heard; or continue crying ad hominums in the wilderness.
The server was a dual-proc machine.
no. read the article again. the server was dual processor capable, but only had a single PIII installed.
Ok well...
... What did that last statement have to do with the whole thing? Absolutely nothing, but the fact still remains that thy were using outadted code in thier comparison, and quite possibly, the Linux machine may hav been misconfigured.
In the first place, I don't believe that the performance could be that much better without some problems in configuration of the software they were using, or lack of support in a specific dirivers area of this. I have a hard time believing this benchmark to be unbiassed, or that the results weren't doctored in some way ( as they were)
I mean, lets be honest with ourselves here, 6.1 is old, and there have been several releases of updated Kernels since, as well as that fact that the 2.4 Kernel is just about ready to rear its head (ugly or not)
My Open Source penny, for what its worth...
m0
No, no, nooo, /etc/shelf.conf?
The shop was setup for windows. The way they served linux was clearly misconfigured! Perhaps a few tweaks in
---
---
"Multiple exclamation marks are a sure sign of a sick mind." (Terry Pratchett)
has anybody noticed that they changed some indepth config params on the linux box? Maybe the reason for performance drop at 20 is an explanation for this. Can somebody explain what they did. -they forced the nic in full duplex -they set the Tag command queuing depth on the SCSI card to 24? -they adjusted the bdflush trigger
It points out some of the things to look out for in this sort of benchmark that can skew the results towards Linux or NT. In particular Linux performs better in the case of the dataset not fitting in RAM whereas NT does better if the data can all be fit into RAM. I'm not sure what the story is with ServerBench.
First of all who would pay 14 million for a box like that??? I know most companies wouldnt and where is the testing software??? Im just wondering where this information is.
The Beaver The Best Things In Life Are Free And So Is Linux!
One linux isnt like windows at all it doesnt need 20 frecking million lines of code to just crash. The kernel is nice and tight, cleanly code and examined all the world over and tested all the world over in REAL WORLD situations this is something microsoft cant touch. And personally I thing linux is a better quality of an OS than windows will ever be. Face it the only reason people use windows is because it was forced down our throats.
The Beaver The Best Things In Life Are Free And So Is Linux!
Maybe what we should do is have an independent test. Linux Journal vs. PCMagazine. Both get the same hardware. Each configures as they see fit their appropriate OS, and then they face off with the same test parameters against each other. Feel free to even make those parameters known beforehand. This is the only way to ever really settle this debate.
If you're rebooting nightly, then you should take a look at your hardware, NOT the OS. I've been running W2K Server for 2 months without a reboot. NT 4.0? I can run the server at least a month without being forced to reboot. If you're installing a lot of crap on your machine (aka shareware, etc) expect it to go down. It's not the OS in question, but rather what you do with it.
I think this is one good thing: the linux OS has forced MS into providing the world with an OS far above what they would have done otherwise. my opinion only.
!#
haha nice one.. anyways, I went to the MS FUD Central and look what I found..
The very definition of Linux as an Open Software effort means that commercial companies like Red Hat will make money by charging for services. Therefore, commercial support services for Linux will be fee-based and will likely be priced at a premium. These costs have to be factored into the total cost model.
And now I am thinking of how much MS is pricing support and service to both private and commercial consumers in addition to the overpriced packages. And also to concider is the fact that their closed environment adds up extra costs to commercial consumers needing to really fix the shit. They often depend on MS. And the depending factor is the oposite of freedom.
People have been fighting for freedom from the date of birth.. This is just another one.. and yes, freedom will prevail sooner or later.
thats exactly what i'm saying. sorry if it got messed up along the way
Upeo
There is always the matter of quality to consider. Linux may slow down, but because of the way it was tested (you know, opensource...) it has a LOT less bugs. i read somewhere that there are at least 1 bug per 100 lines of code. and OS use millions, if not billions of lines. Think about that. WIndows doesn't have the hundreds of thousands of people who have gone in there and seen the bugs for themeselves and fixed iat and told the creators. They only have a limitied team of humans to do the job and humans, as we all know, can only stare at straight code for so long. IT's more of a matter of quality, then speed sometimes.
Upeo
And on another note...+3, Insightful? You're kidding, right?
The real question is this: if Linux had tested better would PC Magazine have published the article? In my opinion it would not. Please be aware that Ziff Davis and Microsoft are essentially married. If you wish to see what the outcome of a test will be simply look at the advertising budget for the two 'competitors' with ZD, and you will be able to predict the results.
Linux will ALWAYS have a disadvantage in these sorts of tests: Microsoft is able to scan the Linux code base looking for cases where Linux is weak and they are not. Once they find a case they call their friends in the press to run a benchmark on exactly the conditions under which Linux will stumble. The Linux community can't do this since 1.) We don't have access to Microsoft's code base. 2). We don't have an 'objective' press in our pocket.
By ZD's tests each version of Windows since 3.1 has been faster than the previous version. That this is a croc is obvious: run Win 3.1 on a simple Pentium 133 with 16 Mb and it flies, run Win 2K on 300 Mhz modern processor with 4 times as much memory and it is sluggish. The trick is that when Microsoft makes a slower OS there is always something in it which is faster. They point out the part that is faster, and that is what 'their' magazines test. The result: the Newer OS is 'faster'. Most people who read the ZD class of magazines can't see through this kind of trickery.
Are the results of the Zdnet bench accurate? Probably; Microsoft set up the test so that it would produce those results, so it is repeatable - just as the Mindcraft test was. Are the tests fair? Of course not, but it doesn't do any good to complain. All that is possible is to fix the problem. When that happens ZD won't re-test: they have achieved their goal of making Microsoft look good to the average person.
This is just another example of Microsoft's dirty tricks department.
I agree with you that ZDnet is now more Windows focussed, but this is entirely due to market share: Windows has a much larger market (and therefore readership) percentage than Linux. A magazine that has more pages on Windows software has more potential buyers than a magazine that has more articles on Linux
However, having a larger number of articles on Windows is not equivalent to being biased. If tomorrow Linux becomes the dominant platform ZDnet will want the Linux market share as well. Therefore, it is not in their interests to alienate the Linux users by posting biased reports.
Philippe
Well gee, just yesterday at Linux.com I read where RedHat Tux(?) outperformed NT by a factor of three. Could be someone should just get some level ground to benchmark.
Having access to the source code of ServerBench would be very helpful though, and right now i cannot test ServerBench because it supports Windows clients only. If anyone from ZDNet listens, i'd love to help porting their client to Linux :-)
Since, when we at slashdot ask for an "unbiased, Non-FUD benchmark" what we really want is something that gives the versimilitude in Linux superiority. Eh, well, here are some links anyway. Take what you want.
First off, some general info (not a benchmark, so it actually contains Truth!)
It may be slightly out of date now that w2k is out, but unix-vs-nt has what appears to be relatively honest benchmarks.
Likewise, PC Week did a review last year shortly after the Mindcraft one
Not a benchmark either, but article on LinuxToday about Linux in the Enterprise.
Each operating system has its own strengths. For example, if I like Microsoft Outlook (for some strange reason), I'd use Windows. But if I like stability, I'd use Linux. For people who want the features that Joe Blow Operating System offers, Joe Blow Operating System is the best choice. And no amount of benchmarks can change that.
Let's stop trying to compare operating systems. Every operating system has its strengths and weaknesses, and subjective standards such as "ease of use" can't be measured through benchmarks. Until the majority of consumers choose Linux, there's no grounds for declaring it the "better" OS.
It says in the article that they noticed how low the results were for Linux, so they tweeked it for better performance.
Most people seem to have sort of glossed this over, with the exception of a few who accuse ZDnet of not tweeking well enough, because they are capitalist pigs who surrender their journalistic integrity to the evil M$ empire.
So I thought I would bring up a question:
Isn't "out of box" kind of significant?
Given, the most hard core of sys-admins should have their boxes tweaked to the gills, but shouldn't your OS perform reasonably well without you having to tweek it?
I know this goes a bit against the whole Linux philosophy, in which anyone worth his salt knows how to max out his own box and has the open source files to do it, but I think out-of-box performance is pretty relevant.
The SPECWeb benchmark didn't mention them tweeking the Win 2k box after they noticed it didn't do as well.
The "problem" here is that Windows has a lot more optimization at the kernel level for transactions and allows almost every process the ability to run at privellaged levels. Linux (Un*x in general as well of course), on the other hand, has a much more secure design in general which prevents many of the stability problems Windows has while sacrificing some performance. This added optimization becomes a factor after about two dozen clients, but once Windows is forced to scale beyond a certain number of clients (100 or so I think), the "optimization" has an inverse effect while Un*x is able to scale with very little noticable difference. This is one of the reasons why NT is often used within firewalls for small to midsized intranet as opposed to larger networks and as web servers.
I believe that the root of this problem lies in the I/O models supported by Linux. Basically, we've got the following:
1) Blocking I/O (used with threads)
2) Nonblocking with select/poll/something else
3) Crappy POSIX aio_* functions (does Linux even support these?)
Windows NT has blocking and nonblocking of course, and has what they call asyncronous sockets that work based on a message queue, but it also has what is called I/O completion ports, which use overlapped I/O. I'm not an expert on these models (yet) but they are the de facto model on NT to support thousands of concurrent connections. It uses a mixture of threads and asyncronous operations (not message queue based this time) so, say, 64 clients are handled by one thread. As apposed to Apache fork()ing whenever a new connection comes in. Yuck.
I've searched for info on making a similar I/O model on linux and have come up with a few references to IOCP on the linux kernel mailing list, but it doesn't seem to have gone anywhere.
If someone could share more information... please do so.
khaladan
Don't know, but the utilities still have the user interface of the Berkeley Unix command line code.
...phil
...phil
"For a list of the ways which technology has failed to improve our quality of life, press 3."
also probably issues with the VM management could be issues here as well. Its well known the the Linux Virtual Memory manager ain't the best. The BSD ones seem to perform the best, but Solaris 8's new algorithm is pretty impressive too.
Is this yet another case where benchmarks have been skewed severely to show a deficiency that doesn't exist? Or is this another area where Linux needs improvement?
We'll never really know, but let's have a purile 400 message discussion* while we don't find out!
* In a very loose sense of the word
-----
I've had a read of the article, and I've read the serverbench page... Now, I may be being a little dense here; might have missed a hyperlink which would have explained it, but:
What sort of client/server are they testing? HTTP? SMB? FTP? SMTP? POP3? I can't see anything which specifies this... and until that's known, nobody can comment on the results.
--
The answer is simple. Many of these tests are just marketing efforts by one company or the other (mostly MSFT) and so the reason for the 'tests' is to attack a competitor or THE current competitor. Today that is Linux. If the Linux kernel v2.4 is really THAT good and I was MSFT (I'm not), I'd get as many 'benchmarks' out now while the 2.2 kernel is all over the place. They are going to have to keep shut or lie once the v2.4 kernel is shipping. I've seen this happen with LanServer 5 years ago and only once was someone dumb enough to compare it with MSFT and NOVELL. If I remember correctly, WarpServer out performed the other 2 with 1 CPU while the others were running 2 CPU's. It was only 'benchmarked' that one time and IBM's marketing deptartment is run be Beavis and Butthead types.
;/
Anyway, it comes down to marketing and Linux IS the competition for the Microsoft Marketing Company....
IMHO
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
Just out of curiosity: Did anyone else get the impression that the RedHat system was configured to handle the SCSI adapter using a loadable module? Is there a performance disadvantage to running your SCSI drives using the loadable module as opposed to having the driver resident in the kernel? I would have thought that they'd rebuild the kernel and include the driver in the kernel.
Also, why not use a Pentium optimized distribution, like Mandrake, instead of the generic 386 oriented RedHat? All these magazine testers seem to do this; apparently when they think of Linux the gears in their heads turn only once and they come up with RedHat (no offense, RH, I use your stuff and am happy with it).
--
CUR ALLOC 20195.....5804M
Although Apache is single-threaded (prior to 2.0, when multithread becomes an option), each request is run in a separate process. If one request stalls, it will still give up its CPU time for other tasks just as nicely as it would under a multithreaded Windows server.
Heh, reminds me of a story I heard about a sys admin of a VAX cluster (iirc). He was lamenting to the DEC tech. support guys that the performance was great until there were about 33 users logged in, then everything would go to hell. He suggested to the DEC guys (in jest) that they go tell the software writers to grep through the code and when they see the number 33, change it to 50. :)
Admit nothing, deny everything and make counter-accusations.
Tweaking is.
If for example Matrox (or anyone else) cheats in their video-drivers, to get better frame rates in Quake,
this could lead to instability, graphical glitches etc.
Why are you making the assumption that ZDnet is
a more serious publication than Linux Journal?
Why do you even make the assumption that ZDnet
is less biased?
ZDnet is a huge publicationcompany that makes about 95% of it's computermagazine-income from
selling windows-mags.
PC Magazine (is it the largest?) is around 99% windows and 1% other. It is obviously much less biased than Linux Today.
Some of their articles are created with the sole purpose of angering Slashdot-readers, thus getting
plenty of readers (because you just have to read
the garbage), and generating income.
>Of course, manufacturers with no morals (ATI, Megabyte, Intel) can optimize for these types of benchmarks, and thus seem faster than they are
No matter what benchmarks exist, _most_ vendors will to some degree optimize to those benchmarks. This is as true of Quake3 or your COM+ example or SPEC or AIM as it is of WinBench3D. Interestingly, the easier it is for the vendors to look at and understand how a benchmark works, the more specific their optimizations will be, and the least subvertible benchmark would be one where the vendors have _no idea_ what it'll do until it's run. I know that open-sourcers aren't going to like that idea, but there it is.
Slashdot - News for Herds. Stuff that Splatters.
Some food for thought: what is a transaction? Their FAQ doesn't seem to cover exactly what it is. If all they're testing is static page serving, it's about as useful a test as seeing how fast it can delete files....
I am bit confused about your WDM ideas. All of Win2k's drivers are WDM. Miniports use a WDM compliant port driver with some new miniport calls. Using "native" nt drivers are what Microsoft calls legacy drivers. The presense of one of these drivers breaks power management on the machine because the legacy driver can not answer the request to change system power state.
If the hal (much smaller in win2k), pci, disk, class, scsiport, ndis, etc. are all WDM compliant. WDM is the native model for Win2k. The comparison to Win32 is not correct. While a Win32 call is really just a wrapper for the native nt call, WDM is usually not a wrapper for legacy calls. It is often the opposite; most legacy functions are just macros that call the new WDM functions.
The main points of WDM on Win2k are cross platform design (and binary compatibility in many cases) and bringing power managment and plug and play to nt. PM and Pnp are considered very important at MS; they are not just political marketing ideas. Although this new model helps the new developer making products for 9x and win2k, it hurts current nt developers by forcing them to rewrite their drivers. In many ways, WDM actually slowed the developement of drivers for Win2k. Finding classes on a moving spec was a difficult task.
My main issue where Win2k is not tuned for performance involve the use of so many general drivers to handle whole classes of devices but none of them well. IMHO MS would have been better off supporting I2O like almost every other OS does. In order to get the best performance, any storage developer will tell you that you have to write a full port drivers to replace scsiport/miniport and disk/class.
NDIS is similar. In fact, it is worse because MS will not certify a non-NDIS network driver. Here, the developer is forced into a slow model. NDIS 5 addressed some issues (removal of some locks, some off load support) but left many others.
It is this push towards a miniport model that I find hurts performance. Also, the messaging scheme for drivers requires system calls between each layer instead of using a direct function pointer interface. Almost all messages must go all the way through the stack, even when most drivers just blindly pass the call down. While this allows any driver or filter on the stack to change things, it slows performance. Many messages must be handled "on the way up" the driver stack. Whole stacks have to wait on events to be triggered and callbacks called. Microsoft gets extendability through the support of upper and lower filter drivers at the expense of performance.
A faster model would reduce the layers in the device stack, use direct function calls rather than system calls for message passing and reduce the need for callbacks and event waiting.
As for reading NT 3.0 documentation...well that's great for discussing NT 3.0. Win2k is their current model and it has changed in ways that are more than just wrapper functions and cosmetic changes.
-- soldack
NT "runs all its servers in kernel space"? Do you mean drivers? Services? Services are not run in kernel space, althougth they can be set with a high priority. All drivers are run in kernel space, though.
"The kernel has a lot of design concessions that faccilitate a really high I/O rate." Really? Have you looked at the code for network and storage? I have written network and storage drivers for NT4/Win2k and is not designed to be fast. Check out the DDK. Both storage and network use a miniport model (SCSI Miniport and NDIS Miniport) with a port driver doing much of the work. To make matters worse, Win2k use WDM for its drivers. WDM tends to add an additional driver object to the layered model. Both miniport and WDM are designed to be very general and take control away from the driver developer. A call to read a few bytes from the disk goes through so many layers. First, the file system drivers, then class.sys, then disk.sys, then scsiport.sys, then vendorscsiminiport.sys, then hardware. There can also be any number of filter drivers in the mix. WDM allows upper and lower filters for each FDO (Functional Device Object). We got a nice performance boost by not using the SCSIMiniport/Class driver interface. Win2k is not designed to be fast as much as extendable and general.
Just my $.02...
-- soldack
TPC benchmarks require the system to be available to be purchased within a certain time frame (6 months?).
The vendors usually use the latest possible software (unless it has performance issues!).
Be Free: Free Software Tuition
Though the machine was single proc, the threading could still have something to do with it. If one of the threads was stalling (a transaction was taking to long, it was blocking on some kernel call, etc) the other threads would be sheduled in their place, and performance wouldn't be so bad. However, in a single threaded design, when one transaction stalls, the whole thing stalls, and thus performance suffers.
A deep unwavering belief is a sure sign you're missing something...
For those who are complaining about them using an "old" version of Linux, get over it. The 2.4 kernel is not only still experimental, it is not in any distro yet. As for the system software being at 6.1, again it doesn't really matter because RedHat 6.2 (The latest from the "Linux company," at least as far as the mainstream is concerned) only contains a few tweeks, and is really not much of an upgrade.
A deep unwavering belief is a sure sign you're missing something...
Okay, now you're into debatable territory here. From my point of view, NT is a microkernel. Yes it runs most services in user space, and the servers in NT are conglomerated into the executive. However, they are still seperate entities, and communicate by message passing. In my opinion, taht constitutes a microkernel.
Having high-performance I/O is a good thing. I think you misunderstand me, I really do like NT, in fact I use it about as often as I use BeOS. I was responding to your comment that NT was designed to be extendible over being high-performance. As for my backing up that NT is tuned more for performance than flexibility, I kind of explained that in the next paragraph, (running services in kernel mode, DirectX in the HAL, etc.) Even if WDM is the "official" driver API for NT, it is still not the native API. Win32 is the official API for NT, however it is not the native API. If you really want to see your applications perform as well as they can you'd use standard NT drivers, and the NT Native API. Neither are sanctioned by MS, but this is a technical discussion, not a political one. The truth is, that MS had to heavily endorse WDM because it needed drivers for NT. Even if it isn't the fastest way to do it on NT, that's the one they had to support. And I think the whole point of the WDM is that drivers should be cross compatible, except for video drivers (which aren't written to the WDM.) I think integrating DirectX into the HAL is a good idea. However, it is quite a complex system for something in the HAL, and does introduce bugs and make the system less "clean" from an acedemic point of view. I was using this point to support my assertion that NT is more tuned for performance then generality. Instead of adding a more general HAL bypass system, they chose to simply allow DirectX to pass through. NT does run a lot of services in kernel mode. In other microkernel OSs, there is no executive. Stuff that is in the NT executive (like I/O managers, etc) traditionally run in user mode. However, by running these in kernel mode, NT gains a performance increase. Also, that is one reason why NT blurs the lines between a monolithic and micro kernel. And I'm pretty sure NT is a microkernel. NT was designed as a microkernel, and its subsystems communicate by passing messages. (If it walks and quacks like a microkernel...) However, the design concessions (for performance) that MS made, made it much more of the "macrokernel" that MS holds it is today. However, if you read the documentation from NT 3.0 or 3.1, you'll see that it is billed as a microkernel.
A deep unwavering belief is a sure sign you're missing something...
NT was designed with a driver architecture. WDM is not it. Even if all NT driver development have shifted over to WDM, it is still not the optimal driver architecture for NT. The architecture that was originally designed to support the original driver API is not easily changed, thus WDM is sort of a bolt-on API for both Win98 and NT. Same thing with DirectX. DirectX is also a bolt-on API to NT, and thus will never be as good as an API around which the NT architecture is designed. Its probably less of an issue for DirectX, because it really just shoves the OS out of the way, but a non-native driver API can cause some performance problems. Maybe the new WDM model is MSs move to make NT a little more general. However, the NT 3.x architecture is still mostly present in Windows 2000. That architecture was designed mostly for performance, thus any discussion of the Windows 2K architecture should still make sense if you talk about the NT 3.x docs. Also, Win32 calls aren't wrappers. Win32 applications make Win32 calls to a user-space Win32 server. The server then communicates with the kernel and kernel level subsystems (via the NT native API and message passing) to implement these calls.
A deep unwavering belief is a sure sign you're missing something...
It's supposed to read
throwing up [h]is cake on IIS
ZZ
In that case, check out the last set of Win2K results which Compaq/MS have withdrawn, presumably because either a) they've been overshadowed or b) they want to put something even faster up. Effectively, at this moment in time, Win2K has the top three performance slots and the top ten price/performance slots.
--
Cheers
Cheers
Jon
I find this rather odd given this story from a couple of days ago. Maybe the SPEC group knows a thing or two more about setting up Linux than the PC Magazine guys do. Or maybe the PC Magazine guys know a thing or two more about setting up Windows. Or maybe a bit of both -- PC Magazine's traditionally been a Windows shop so I'd expect their Windows know-how to be much more advanced. I don't know anything about the SPEC group, though.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
This just went up on the TPC website Monday, there is a monster leader in transaction processing price/performance and that is:
IBM Netfinity with Intel Xeon processors
IBM DB2
and Windows 2000.
You will not believe this unless you see it!
Read'em and weep
"a powerful and unexpected ally..."
Am I allowed to laugh about this? :)
If we just give the advocates/zealots/other craptalking people a separate forum, we can then go on with talking about Stuff That Matters(tm) over HERE. Thank you.
--
Never underestimate the relief of true separation of Religion and State.
Anyone else notice that NT 4 performed better than Win2k? Hmmm, maybe microsoft will start learning the reason why things like VI are still around since the seventies in any *NIX distro: They Work!
Newer does not always mean better, but it does mean that you lack all the special features
Lemure, wtf! Don't you mean Lemur?
View Here
Eh...
Where the hell does this drop at the number 20 come from? It's like, good...good...good, DOH! 20, BOMB OUT! It's not even a power of 2! *VBG*
This really makes little sense. Also, what about other Linux distros, is RedHat optimized in some way that causes this?
I can hardly see why Linux would just go apeshit once 20 users is met. I mean, you think that you'd see more of a slight downward progression than the floor falling out.
Eh...
They were serving applications...
Eh...
They said they used Linux 6.1 in a couple of places....... Since its an odd numbered version, I guess that means its the experimental branch. Still, their time travlleing skill are quite impressive.
I think a couple of things need to be specificly pointed out here. First off, with this particular benchmark, its indicated that a single 'client' actually represents several dozen users in real life. So, RH 6.1 Pro takes up to 20 'clients' before its score dips, I'd certainly like to know what that equates to in real life myself. Perhaps there's just a small bug in Red Hat that reduces its performance in what appears to be a very synthetic benchmark. Secondly, the testers who performed the benchmarks didn't appear to know what they were doing to well. Can we be sure that they truely had their Linux box customized and tricked out for these tests? No, of course not. Third, don't discount the possibility that these scores may be entirely accurate. Even a broken clock is right twice a day. -Algernon
If you find this hard to believe, check out these mind blowing statistics. Who would have ever thought?!
God Fucking Damnit
First, let me say, that benchmarks like this are useless. I'm not against benchmarking, mind you, but what I AM against is artificialy benchmarks. For example, in testing 3D cards, ZDN still uses an artificial benchmark called WinBench3D. Of course, manufacturers with no morals (ATI, Megabyte, Intel) can optimize for these types of benchmarks, and thus seem faster than they are. However, if you do a real world benchmark, like say test the FPS in Quake, you're results are actually valid. If a manufacturer cheats so their card runs actual games faster, then thats actually a good thing. What I'd like to see is a real world server benchmark. Maybe set up a COM+ simulation where actual COM+ applications (for example a database) does actual requests on the server, then measure how many clients the server can handle. Or do something like have the server serve up a database and have scripted clients access the database in a real manner. Those are the kinds of benchmarks that really work, but unfortuenatly, they take actual work. As for NT beating Linux, remember that NT is designed to run stuff like this. Though NT is a microkernel, it rusn all its servers in kernel space (though Linux does as well, I think). Also, the kernel has a lot of design concessions that faccilitate a really high I/O rate. There not so good for doing real world tasks, because the OS tends to step on its own toes, but if you testing raw performance, NT usually wins. But these performance enhancements take their toll on stability and ability to handle high loads. For example, in Windows2K, DirectX has some interface calls implemented into the hardware abstraction layer, which really speeds up performance, but at the cost of stability. That said, WindowsNT really IS a decent OS, and some parts are simply better designed than their conterparts in Linux. For the desktop, (if you have the RAM), W2K is perfectly stable (because most desktop users reboot at least once every few days) and nicely supports media. Also, Win2K has a good multi-threaded TCP/IP stack that was rewritten from NT4. Despite its faults (ahem bloat) it does actually have some features that the Linux guys would be wise to look into. (ahem, COM) W2K is nowhere near being the end-all be all of OSs, but neither is Linux. They both have their flaws, and NT has actually improved enough that W2K actually has some uses in the server role! Anything doing small transactions that doesn't need to be particularly stable (for example DNS) would be served well by NT which responds well to little transactions like this.
A deep unwavering belief is a sure sign you're missing something...
That could also be a problem. Processes take more time to start than a thread does. Also, it still doesn't mitigate the fact that the TCP/IP stack is single threaded so THAT can stall.
A deep unwavering belief is a sure sign you're missing something...
This whole bench test is pretty useless. Ziff wanted an "application" benchmark that was cross-platform and didn't rely on applications. What they actually built was so content-free that it simply tests network and OS performance as far as the TCP/IP stack.
Not surprisingly, they found that the Win TCP stack is quicker than the (known to be single-threaded) Linux stack. QEFD.
I'd like to see better benchmarks, but I'd much rather see something for simple Corba vs. Corba, or Corba vs DCOM. SOAP (the Apache approach of deployable handlers), vs. SOAP (Servlets) vs. SOAP (Microsft's SOAP-on-a-ROPE) would be even more interesting. We're doing something along those lines ourselves - maybe it will be publically publishable.
To get the alternative "Useless benchmark shows Linux to be faster than Windows" story go here.
Why is it always Linux vs. Windows?
Because that's what Linux advocates trump up. Ever since Linux became 'popular', advocates have been pitting it against the big bad evil Microsoft. Nevermind that until recently, Solaris was just as closed-source and dealt in the same underhanded tricks as Microsoft. Nevermind that they're two completely different types of operating systems aimed at two entirely different classes of people.
Basically, Linux people want Linux to be able to do everything that Windows can. They want it to be a robust server operating system. They want it to be an easy-to-use client operating system. They want it to run everything. They want to be the monopoly (but a monopoly of choice, not of force). Nevermind that Windows 2000 isn't trumped as the OS for everyone and Windows 98 isn't used in high-end server systems (and yet, advocates want Linux to do all of these tasks, and rule the hand-held market as well). And so, we get tests like this, Win2K vs. Linux, when really, what we should be getting is Win2K vs. Solaris (which I'm quite confident would blow Win2K out of the water).
Does Linux really want to compete at the levels of AIX and Solaris?
No, they want to compete with Windows. Windows is the enemy. Sound the alarms, and when Windows does something better than Linux, something is seriously wrong with the world (or so they would have you believe). Perhaps what would be a better suite of tests for Linux is one which isn't a comparison test at all, but rather one which looks for deficiencies so that people can start fixing them and quit debating about whether or not a comparison is valid.
Because that's what Linux advocates trump up. Ever since Linux became 'popular', advocates have been pitting it against the big bad evil Microsoft. Nevermind that until recently, Solaris was just as closed-source and dealt in the same underhanded tricks as Microsoft. Nevermind that they're two completely different types of operating systems aimed at two entirely different classes of people.
Perhaps a less biased way of saying this is "Because Windows is, arguably, the main competition for Linux. While AIX & Solaris are also viewed as competitors, due to Linux' current weakness in scalability, they are not considered direct competitors."
Now, that said, I'll respond by saying you're an idiot. Linux & Windows 2k ARE NOT designed for two different types of users. Both are designed for general use, high-end workstations low-to-mid end servers. In particular, in the context of the question, they are designed for EXACTLY the same market.
As far as AIX & Solaris, they are also the competition. But, most people who have the budget to run a high -end unix server have a reason to spend the money (Support, a boss that's an idiot, or a need for specialized capabilities or scalability that Linux & Windows don't allow). Linux is rapidly advancing, & is beginning to address the last two issues (scalability & features), but at present it's hard to directly compare Linux to some of the commercial Unixes. And of course, you again need to consider the context. Since the question was specifically in response to a benchmark comparing Linux to Win2k, why would you even expect AIX or Solaris to be brought up?
diff -u linux/net/ipv4/tcp.c:1.1.1.6
@@ -1575,7 +1575,7 @@
add_wait_queue(sk->sleep, &wait);
for (;;) {
- current->state = TASK_INTERRUPTIBLE;
+ current->state = TASK_INTERRUPTIBLE | TASK_WAKE_ONE;
Offhand, it looks like that particular change isn't in Red Hat 6.1 or 6.2. I don't know whether this would affect ServerBench performance, though. It's hard to tell without looking at the source.
The server was a dual-proc machine. Win2k has a multi-threaded TCP/IP stack, linux 2.2.x doesn't. That probably accounts for most of the issue right there - at around 24 users, the single processor limitation of the Linux TCP/IP stack was reached, and the Win2k mahcine just split the load up.
Of course, IANALOLT (I am not Alan Cox or Linus Torvalds), but it seems the most likely explanation to me...
---- I made the Kessel Run in under 11 parsecs.
This just went up on the TPC website Monday, there is a monster leader in transaction processing price/performance and that is:
You will not believe this unless you see it!
Yes - but check out the hardware. 32 four-way pentium Xeon's, and over a terabyte of disc space, and an obscene amount of RAM. That is not a standard setup, although it was built with standard parts (trust me - I know the team which built it). That is not to say that the DB2 team isn't extremely pleased with this result :-)
Just because it's running on Windows 2000 does not automatically mean that there might not be better choices for an OS to support this benchmark. It's not even entirely clear to me that Windows NT might not have been faster here, given the benchmarks which MS put out on their own website showing that Windows 2000 does better in limited memory, but is worse than NT above 128MB (and these machines had a lot more than that). Remember that DB2 UDB has a shared-nothing architecture which that it scales extremely well and is additionally capable of using raw devices so the OS in question may not have a big impact on performance. And DB2 runs on most platforms out there, from OS/2, AIX, HP-UX, Solaris, Linux, Windows 9x/NT/2000, SGI, SCO, Dynix and various 64 bit platforms as well.
Of course, it would be nice to have some side-by-side benchmarks of DB2 UDB on Windows NT/2000 and DB2 UDB on Linux. There will almost certainly be some benchmarks on Linux sooner or later - since IBM has made Linux available for all its machines, it makes sense to publicise the performance of its flagship DB product on Linux as well.
Cheers,
Toby Haynes
P.S. I work on DB2 UDB development.
Anything I post is strictly my own thoughts and doesn't necessarily have anything to do with the opinions of IBM.
"Each was tested on it's own network, with 2 subnets of 24 servers, the windows network consisted of 48 PII's, whereas the linux network had the added advantage of having a Cray Supercomputer making requests at full charge on the 24th node of the first subnet..."
Eh...
I hate people talking how 2.4 will fix everything, 2.2 surely didn't.
Where did I say that 2.4 will fix everything?
I said that there is a specific problem, known in 2.2 that has turned up before, that is a potential explanation for this bad result.
There are other known (and fixed) scheduler problems.
Encountering any combination of these in 2.2 benchmarks is to be expected. Don't make these out to be more or less than indications that 2.2 had some obvious room for improvement.
I am sure that 2.4 will have more problems. However many problems that turned up in benchmarking 2.2 have been fixed (because they turned up in benchmarking 2.2), and preliminary benchmarks of 2.4 (eg the recent SpecWeb result where it nearly tripled Windows 2000 on a similar 4 CPU box) indicate this.
Now will 2.4 be ready for the enterprise, as they like to say? Not really. First of all until it has been through a few point releases, I would expect some significant bugs. (To be expected in any software.) Aside from that issue, it lacks many managability tools, a volume manager, more work needs to be done on failover, journaling filesystems are needed, etc. I have been convinced by Larry McVoy's argument that further work on SMP is not needed, NUMA (done through clustering and virtual operating systems) is.
These are known problems. Work is being done on them. However there will be room for complaint about Linux vs more mature systems for some time to come. However problems are getting solved, and Linux is moving up the food chain, fast.
Regards,
Ben
My usual seat in the cluetrain is at A HREF="http://pub4.ezboard.com/biwethey.ht
An immediate thought.
The "thundering herd" problem that was identified in Mindcraft and fixed in 2.4, isn't that still present in RedHat 6.1? (BTW calling it "Linux 6.1" really irritated me.) That could explain a sudden drop-off. It is not a problem, not a problem, then suddenly becomes a problem and as soon as you get a slow-down, you get a real traffic jam.
Just guessing...
Ben
My usual seat in the cluetrain is at A HREF="http://pub4.ezboard.com/biwethey.ht
I know this sounds strange but when I'm looking at designing a high transaction application or site I don't even LOOK at Windows or Linux. Does it suprise me that Linux doesn't scale to the enterprise market ? No, its written by individuals for lowish demand systems that they require, rather than by Company A who is implementing for Company B something that will cost several million pounds of development.
These sort of tests are IMO unfair to Linux. Should you use NT/W2K or Linux for your high transaction application/site ? The choice is more normally "Should I use True64, AIX or Solaris ?".
Linux works great for me as a webserver, as a client who takes a limited number of hits at a cheap price. If you want to scale you buy more boxes.
On the back end use a large end server with lots of RAM that has a massive IO throughput.
Does Linux really want to compete at the levels of AIX and Solaris ? Why not go for the niche, of cheap, reliable, and easy to scale horizontally.
An Eye for an Eye will make the whole world blind - Gandhi
Okay, they set the ethercard to full duplex and increased the queue depth on the scsi card. Fine -- makes sense. Stopped the "atime" updates. Makes sense.
... interesting. We're developing a new filesystem, and ended up ignoring bdflush completely to get good performance. Here's what those values mean:
/usr/src/linux/Documentation/filesystems/proc.txt:
/proc/sys/vm/bdflush
... they seem to have changed one of the "dummy" values... wonder why? Other than that, they appear to have increased the interval at which bdflush runs, meaning more stuff is hanging around in memory. It may be that at 24 clients, bdflush is banging on the filesystem too much. I would loveto see a graph of disk activity inluded with the results. Sometimes Linux will go through a silent-storm-silent-storm cycle as bdflush runs on a busy system. It would be interesting to see how a journaled filesystem would perform. I think Reiser does his own buffer-flushing rather than relying on bdflush runs to do it, meaning he has finer control over it. It would also be interesting to see this test run on FreeBSD, which does a better job keeping the disks busy.
/usr/src/linux/Documentation/IRQ-affinity.txt ... describes how to have specific CPUs handle specific IRQs -- like the mindcraft tests did with NT.
But they also did this:
echo 100 5000 640 2560 150 30000 5000 1884 2 >/proc/sys/vm/bdflush
From
Table 2-2: Parameters in
Value (default/tweaked)
nfract (40/100)
Percentage of buffer cache dirty to activate bdflush
ndirty (500/5000)
Maximum number of dirty blocks to write out per wake-cycle
nrefill (64/640)
Number of clean buffers to try to obtain each time we call refill
nref_dirt (256/2560)
buffer threshold for activating bdflush when trying to refill buffers.
dummy (500/150)
Unused
age_buffer (3000/30000)
Time for normal buffer to age before we flush it
age_super (500/5000)
Time for superblock to age before we flush it
dummy (1884/1884)
Unused
dummy (2/2)
Unused
Tweakers may also be interested in reading
ServerBench is not available in source code, and the testing was done by ZDNet. From what i know about ServerBench it uses a threaded IO model on NT, but a fork/process model on Linux. The Linux 'solution' is coded by ZDNet, with no possibility from us to influence/comment the design and approach used at all. Even under these circumstances we expect the 2.4 Linux kernel to perform significantly better in ServerBench than 2.2 kernels. The 2.2.1x (and late 2.3.x) kernels had some VM problems, and with increasing VM utilization (more clients) this problem could have been triggered.
SPECweb99 OTOH is a standardized benchmark with full source-code access (ServerBench are closed binaries), so all SPECweb99 implementational details are visible.
Nevertheless it's technically possible that ServerBench triggers performance bugs in Linux - we'd love to see the source to fix those bugs ASAP, if they are still present in 2.4.