FreeBSD 8.0 vs. Ubuntu 9.10 Benchmarks

An anonymous reader writes "Phoronix has brought benchmarks comparing the FreeBSD 8.0-RC and Ubuntu 9.10 Alpha 6 operating systems. FreeBSD rather ends up taking a wallop to Ubuntu Linux, but there are a few areas where FreeBSD 8 ran well. They also posted benchmarks comparing this near-final FreeBSD 8.0 build to that of FreeBSD 7.2 to show performance improvements there but with a few regressions."

Comments

Let's see these against my Gentoo...

[Jurily]

...once I'm done compiling.

Re:Let's see these against my Gentoo...

[wall0159]

Given that, I think it'd be fairer to compare Gentoo against Ubuntu 10.04, don't you?

Re:Let's see these against my Gentoo...

[TheBlackMan]

...once I'm done compiling.

If they used stock kernel (and they surely did), then any optimized Linux will be faster than that. As we can see, even "bloated" Linux is still pretty damn fast.

What's the point.

[hedwards]

I'm sort of curious what the point is of comparing an alpha to a release candidate. Additionally it's a minor update versus a major update. Throwing in an older release makes it all the more pointless as I'm not seeing anywhere in the summary that they disabled debugging.

Re:What's the point.

[bostei2008]

Just a plug for the Phoronix Test Suite?

Re:What's the point.

[Jurily]

I'm sort of curious what the point is of comparing an alpha to a release candidate. Additionally it's a minor update versus a major update. Throwing in an older release makes it all the more pointless as I'm not seeing anywhere in the summary that they disabled debugging.

They left out almost all distros, too.

Re:What's the point.

[hedwards]

Well, with the number of distros out there, it's inevitable that they'd leave out most of them. But I must admit that leaving out every other distro makes it kind of hard to know if Ubuntu is doing better than the pack or about the same.

It's kind of suspect, in my opinion, that the older release was doing so much better than the newer one, considering all the time that's been spent in recent times on optimizing various portions of the source. It's also worth noting that probably a much larger portion of the FreeBSD user base will recompile their kernel pretty much immediately with basic optimizations and removing the cruft that they don't need or want.

Re:What's the point.

[fbjon]

Is there any actual benefit to be gained from removing "cruft", other than saving a smidgen of memory?

Re:What's the point.

[suso]

Yes and we can't have advertising on the net! *sarc*

Re:What's the point.

[jhol13]

The comparision might have made sense if they had tested whether the different version of the compiler used is the "culprit" or not.

Re:What's the point.

[geminidomino]

Is there any actual benefit to be gained from removing "cruft", other than saving a smidgen of memory?

Long, but not long enough answer:

Performance: Unless the cruft is a bunch of data or NOPs, it will be executed at some point, which is pointless (or it wouldn't be cruft.) And whether it's data or instruction, if "good" data is swapped out of the cache in favor of the cruft, then it will have to be read back in (cache misses).

Security: Bugs love to hide in cruft.

tl;dr version: Yes.

Re:What's the point.

[greed]

Yes.

Any resource allocated is something that needs to be tracked. Those tracking objects take a little bit of management time. What kind of time and when depends on the object: sleeping processes aren't checked by the scheduler, but other process operations like signal need to.

Directory entries add up. Even if you have a hash or tree directory index structure, you may still have to scan the whole thing to find files by pattern (for example). With many directory indexes, you must do an end-to-end scan to find out if a name is truly unique.

And almost ALL of the "cruft" that is added to new systems is stuff that "does things" automatically. In order to do that, they all wake up on certain events (including timer events) and check to see if they need to do something. Like multiple devices sharing an interrupt, it all adds up.

Sure, a new system has enough power that it retains good interactive response with all that. But run an old OS on a new system, and wow is it fast.

Re:What's the point.

[Ritz_Just_Ritz]

Thinly disguised? About as "thinly disguised" as Divine's "woman costume." :)

Re:What's the point.

[kad77]

I'm sort of curious why the slashdot story summary is so annoyingly biased in it's phrasing - "FreeBSD rather ends up taking a wallop to Ubuntu Linux, but there are a few areas where FreeBSD 8 ran well", when the arguably flawed test suite shows NO SUCH THING!

The FreeBSD system has very comparable or better benchmarks on nearly every metric in the test, just click through TFA and see for yourself.

Tripe.

Besides the needless and counterproductive bias, the phrase X "rather ends up taking a wallop to" Y is clunky and sophomoric. Editors, get a life.

Re:What's the point.

[Abreu]

For want of a nail the shoe was lost.
For want of a shoe the horse was lost.
For want of a horse the rider was lost.
For want of a rider the battle was lost.
For want of a battle the kingdom was lost.
And all for the want of a horseshoe nail.

Re:What's the point.

[WinterSolstice]

I must say that I'm really fed up with these sorts of benchmarks. They insult the intelligence of IT/technical people AND business people/endusers.

I need to find or create a site that actually benchmarks a variety of systems (windows/linux/unix/bsd) for ACTUAL scenarios that the users of the systems expect them to perform. Not sure how important LAME encoding is for BSD people - I know I don't use it for that. Not sure how many Ubuntu users run DB servers or firewall routers.

Can anyone point me to someplace like this? Someplace where I can see, for example, DB driven dynamic pages served per second - concurrent DB users, scalability to 4+ processors, etc?

I want metrics like SAN performance, DB performance, mail server performance, HA, clustering, firewall performance etc for server "distros" (Windows Server, AIX, Solaris, BSD, Linux, OSX Server)
I want metrics like graphics benchmarks, app response time, those classic GUI task tests for desktop "distros" (Windows, OSX, Linux, BSD).

It annoys me that I see benchmarks that are so narrow and pointless. I can't make reasonable non-fanboy type comparisons when the benchmark doesn't include anything that matters to me.

Re:What's the point.

[compass46]

I don't believe commenting out a bunch of device drivers from GENERIC for hardware that is not even in the box will have much of an overall effect really. Yes, there are a handful of other options in GENERIC but how many of those do ppl actually comment out and how much of an effect is there really?

Re:What's the point.

[Anonymous Coward]

True. And let's also see some benchmarks with the system under heavy load. Run it as a webserver with thousands of page requests a second. Run it as a database server under a realistic load (not simply inserting a few records).

I could give a rats ass about how fast it compiles some app, show me some REAL benchmark!

Re:What's the point.

[MollyB]

I was hoping someone would start a technical discussion of dependencies...

Re:What's the point.

[WinterSolstice]

Thank you. That's my point exactly.
Benchmark testing is very unscientific it would seem, since there is rarely ever a "control" and the test suites seem to run nearly random load sets.

They're also almost always impossible to reproduce since the test suites are typically quite expensive... and many companies forbid publishing benchmarks!

It'd be nice if a 3rd party like W3C did a webserver comparison where every company got to bring their big guns (each server configured by the *vendor* for best performance on given hardware) and then the W3C put it up on a standard heavy load tester than ramped all the way up to "crushing load".

Same with databases.

I want to see not just how a server from vendor X would perform, I want to see where it dies.

Re:What's the point.

[BitZtream]

Very little effect. The drivers are generally detached from the kernel if not used. Most things are loadable modules now anyway unless they are core components or so generic that EVERY PC has them.

Re:What's the point.

[MadAhab]

That's about all there is to it.

Re:What's the point.

[Sancho]

I don't understand. Did we follow different links? You have to pay attention to the "more/fewer is better" line--for a lot of those charts, larger bar graphs meant worse performance.

Re:What's the point.

[node 3]

I'm sort of curious what the point is of comparing an alpha to a release candidate.

Alpha is sort of a misnomer. Well, actually, it's extremely accurate for what the term really means, but for the way people think of beta and alpha, Ubuntu's "alpha" is more akin to other organizations' "beta".

Ubuntu 9.10 (for which Alpha 6 is the current version) is due out in a month.

Re:What's the point.

[nwmcsween]

I want none of this, I want actual benchmarks that benchmark the *SYSTEMS* and not an aggregate of every single system.. I want to know scheduler throughput and latency, file system overhead, etc basicly I want deep statistics that show how well the kernels perform and not how some randomly written application performs.

Re:What's the point.

[quickOnTheUptake]

Yeah, but it's phoronix, they do that shit consistently. For example here: Benchmark of OSX and Ubuntu, using compilation of PHP and ImageMagick, but with different versions of GCC.
Honestly, all they do is vomit random statistics onto ad laden pages, and we slashdotters keep taking the bait, hoping somewhere in the noise there will be a couple meaningful numbers.

Re:What's the point.

[mtippett]

Phoronix Test Suite (http://www.phoronix-test-suite.com/) is open source.

Join the community, help define meaningful tests and benchmark.

Re:What's the point.

[WinterSolstice]

Because of exactly the reasons you point out. OSX, Windows, and BSD are considered operating systems, while Linux is not an operating system but a kernel around which a variety of software distributions exist.

I'm pointing out that the nomenclature 'distro' doesn't quite adequately describe the correct term to characterize these - but I'm loosely and generically applying it for the sake of brevity.

Re:What's the point.

[True Grit]

I could give a rats ass about how fast it compiles some app, show me some REAL benchmark!

Compile speed *is* a real benchmark, just obviously not one that interests you, but it is useful to others.

The problem Phoronix has is in finding a set of benchmarks that is useful to the largest segment of their readership. Judging by these sub-threads here, its clearly not easy.

Re:What's the point.

[Eil]

I think you don't understand what a benchmark is. A benchmark is meant to give you a rough idea of how certain (yes, narrow!) tasks perform on a given OS or hardware. Some benchmarks focus on disk-bound tasks, some on memory-bound tasks, some on CPU-bound tasks (which can be further divided into primarily integer or floating-point tasks), or some combination of these. The reason they include so many is because you're supposed to know enough to pick one that roughly matches the application that you care most about. Or take the composite of two or more benchmarks together. For example, for most database applications, you will probably care most about disk I/O performance unless you have some crazy nutty queries and then you'll care about memory and CPU too.

Further, the point of the benchmarks is not to give you absolute numbers for these tasks. Nobody really cares exactly how many seconds it takes Ubuntu to encode an MP3 file, because it's going to be different for every system, even those with ostensibly identical hardware. The information is in the comparison between operating systems. In this case, Ubuntu was found to perform better in most cases than two versions of FreeBSD. That was what the author of the article set out to find and report.

What irks me the most is that your whole complaint is that the benchmark suite doesn't include things that you think are useful. And because the article didn't include those things that you want, it's worthless for everyone, right? Well, the good news is that the Phoronix Test Suite is open source, so you can add whatever tests that you would like and contribute them back to the community. Shall I let them know you'll be sending in your patches soon?

Re:What's the point.

[True Grit]

I don't believe commenting out a bunch of device drivers from GENERIC for hardware that is not even in the box will have much of an overall effect really.

Leaving out drivers for stuff you don't need does have a very big effect on kernel *compile* times, note that the OP referred to *recompiling* the kernel.

Why waste time compiling something you'll never need, and risk the possibility of compile failures from problems in that "cruft" code?

Re:What's the point.

[gowen]

the phrase X "rather ends up taking a wallop to" Y

Furthermore, its pretty hard to figure out whether it means "X beats Y", or "Y beats X". "Taking a wallop to" is a construction similar "Gives a beating to...", but also "Takes a beating from..." From the surrounding text it seems to mean the latter, but I'm not too sure.

Re:What's the point.

[Ofloo]

i think bsd once did a performance test like that comparing several sql daemons including mysqld to linux kernel and not some distro, also an other note freebsd 8 is still development release this isn't a production release, .. while ubuntu 9.10 is, .. for me personally FreeBSD is kickass server os while ubuntu is a nice desktop. Some people might not agree but linux to me has still a long way to go before it is suited to be a good server, then i'm only thinking differences between distros the hier is different on every linux, .. some are simular but they've always got some personal touch, which sucks .. an other thing is still a lot of new things are introduced to linux which also makes it a bit unstable/unsecure sometimes, .. this is good for desktop users but not for servers, which need uptime and need to reboot as little as possible. Knife always cuts both ways new things maybe better performance, .. but this comes at a cost, with every new line of code there is a risk of breaking some code or insecure code might be added. I'm not pro anything but at this point linux is a verry nice desktop, at least to me.

Careful manipulation of readership

[Anonymous Coward]

It's interesting how on page 1, there are three graphs. Two of these are "lower is better" (where Ubuntu wins), however, when FreeBSD wins the graph is displayed in MIPS where "higher is better", thus appearing to make Ubuntu win there too.

If you're a casual reader not paying attention, reading, or clicking on to page 2 (and you know I'm right when I say that's most of the people reading this article), you can see where this is going.

Summary

[__aaxwdb6741]

The article is of little use other than tell the general populace of Slashdot that FreeBSD 8.0 and Ubuntu 9.10 are right around the corner, and that we should be hyped. Also FreeBSD 8 is a little faster than FreeBSD 7.2 but a lot slower than Ubuntu Linux 9.10

I'm not surprised, however I do belong to the group that does not really care about relative performance to other OS's as performance is only one of the aspects from the vector of decisions we had to make to finally choose FreeBSD for mass-scale deployment.

Re:Summary

[mi]

Also FreeBSD 8 is a little faster than FreeBSD 7.2 but a lot slower than Ubuntu Linux 9.10

Not disputing the conclusions of the memory, I/O benchmakrs, part of what they benchmarked is the compiler — FreeBSD's gcc-4.2.1 vs. Ubuntu's 4.4.1. I'm not surprised, GCC's newer release both compiles faster and produces faster binaries.

You could say, FreeBSD is at fault (and thus deserves bad rep) for including an outdated compiler, but, on the other hand, FreeBSD's choice may prove wiser, when bugs in the hot new compiler surface... The compiler's history — as, really, that of any sufficiently complex piece of software — teaches, it is wise to stay several releases behind. Bugs in the older releases are known and all platforms — certainly including FreeBSD — merge fixes into their repositories.

Their choice of using ImageMagick as a test is particularly suspect — that software has so many options (which graphics back-ends to include, whether to use OpenMP, etc.) and varies so greatly between its own frequent minor releases, that I'm sure they built it with subtle (but timing-affecting) differences between platforms... For just one example, FreeBSD's port of ImageMagick runs all of their bundled self-tests after compilation by default, which takes quite a while. Unfortunately, the testers don't even mention, how exactly they built the stuff. If they used the port on FreeBSD, did they change any options? If they did not use the port, then they didn't build ImageMagick the way the users will be building it... And if they did use the port and flipped some features, did they ensure an identical match between two FreeBSD versions and Ubuntu?

Re:Summary

[TheRaven64]

FreeBSD ships with GCC 4.2.1 as the system compiler because it was the last release to be GPLv2 and GPLv3 stuff is not allowed in the base system. GCC 4.4 is in ports, and you can use this to compile ports easily by just setting a flag in make.conf if you care. FreeBSD 9 will hopefully be using LLVM/Clang as the system compiler, which should give it a nice boost.

Phoronix has a history of doing long and misleading benchmarks between Linux and *BSD/Solaris, where they manage to include so many extraneous factors that the results are meaningless.

Re:Summary

[nurb432]

Stability and predictability are also rather important factors ( in the favor of *bsd )

Re:Summary

[Nevyn]

Phoronix has a history of doing long and misleading benchmarks between Linux and *BSD/Solaris, where they manage to include so many extraneous factors that the results are meaningless.

As the saying goes "don't attribute to malice ...".

Phoronix has a history of doing long and misleading benchmarks between Linux and Linux. Phoronix are idiots, if only they wouldn't keep hitting the /. front page *sigh*.

Re:Summary

[TheRaven64]

Phoronix are idiots

That's a bit harsh. They often run some good articles about X.org and related things. It's just when they stray from the area that they actually know about - open source graphics - and go anywhere near benchmarks that they become incompetent.

Benchmarks...

[coryking]

Benchmarks are useless. There are way more important things to judge an operating system then "speed".

Does Ubuntu have nearly as good of documentation? No. It has that "info" nonsense.

Does Ubuntu provide a stable platform to build a server? No. It, like most linux distros, changes whole versions during updates. That isn't stable.

Does Ubuntu provide a way to strip itself down to the bare metal? Ain't as easy as the BSD's.

Is Ubuntu built around solid engineering and design, or politics? Depends--Ubuntu seems to be less afraid of the big bad FSF as other distros, but it still is steeped in an OS built for politics. FreeBSD is pretty tame and tends to focus on solid engineering rather than political maneuvering.

But really, Comparing FreeBSD to Ubuntu is like comparing OpenSolaris to Windows 7. FreeBSD is largely a server operating system were as Ubuntu is an end user operating system. And if you are comparing server operating systems, there are far more important criteria than "speed". Things like version stability are vastly more important.

Re:Benchmarks...

[nxtw]

Does Ubuntu provide a stable platform to build a server? No. It, like most linux distros, changes whole versions during updates. That isn't stable.

Ubuntu has LTS (long-term support) releases which are supported for 5 years on the server side. The last was 8.04 and the next will be 10.04.

I prefer RHEL/CentOS, however. I wonder how many people use Ubuntu LTS instead of using RHEL or SLES instead.

Does Ubuntu provide a way to strip itself down to the bare metal? Ain't as easy as the BSD's.

How often is this important? I can think only of a few situations, such as when fitting a system into a small/cheap flash.

But really, Comparing FreeBSD to Ubuntu is like comparing OpenSolaris to Windows 7. FreeBSD is largely a server operating system were as Ubuntu is an end user operating system. And if you are comparing server operating systems, there are far more important criteria than "speed". Things like version stability are vastly more important.

Ubuntu has a separate 'server' version (which really just includes a different set of packages and a different kernel build.)

Re:Benchmarks...

[agnosticnixie]

1. Agreed, I think neither is really better and use depends on what you need, which of course some Linux and BSD zealots seem to disagree on - apparently the main free *nix are now entrenched enough to be part of the unix holy wars.

2. Varies, FreeBSD doesn't have perfect docs either and other distros do have better docs

3. Lrn2LTR

4. Yes, it does, there's a 10MB barebones installer for the dedicated and if you need less, pick another distro, even FreeBSD will probably be tons of tweaking at this point.

5. Political strawman yay

6. Call the whambulance.

Re:Benchmarks...

[moon3]

Speed is the most important factor for web-server scenarios, if FreeBSD can handle 10x more inserts into SQL-Lite then Ubuntu in the same benchmark, on the same hardware then Ubuntu is KO'ed by BSD in the server arena, no offense.

Re:Benchmarks...

[whatajoke]

Benchmarks are useless. There are way more important things to judge an operating system then "speed".

Does Ubuntu have nearly as good of documentation? No. It has that "info" nonsense.

Does documentation displayed in info contain less content somehow?

Does Ubuntu provide a stable platform to build a server? No. It, like most linux distros, changes whole versions during updates. That isn't stable.

The last LTS release of ubuntu was 8.04. What major version changes have happened to it?

Does Ubuntu provide a way to strip itself down to the bare metal? Ain't as easy as the BSD's.

Bare metal means what? Ubuntu already runs on netbooks. Are you trying to run a mail server on your wrist-watch?

Is Ubuntu built around solid engineering and design, or politics? Depends--Ubuntu seems to be less afraid of the big bad FSF as other distros, but it still is steeped in an OS built for politics. FreeBSD is pretty tame and tends to focus on solid engineering rather than political maneuvering.

Ubuntu is built upon debian. And so inherits a very solid release process from it. Your rhetoric about distros being afraid of FSF is mind bogglingly idiotic. And not worth discussing.

But really, Comparing FreeBSD to Ubuntu is like comparing OpenSolaris to Windows 7. FreeBSD is largely a server operating system were as Ubuntu is an end user operating system. And if you are comparing server operating systems, there are far more important criteria than "speed". Things like version stability are vastly more important.

And how are LTS releases of ubuntu lacking in stability? All in all; do you have any positive feedback for the ubuntu communtiy?

Re:Benchmarks...

[Murdoch5]

Speed is the most important factor hands down, documentation is no where near as important. Second is the ability to configure the application by stripping symbols and optimizing using flags and third is over all memory and cpu foot print. You want it to use the least amount of cycles per operation possible. Then after all of that has been considered, then you look at documentation.

Re:Benchmarks...

[Rogerborg]

Ubuntu [is] steeped in an OS built for politics

Hah, I see what you did there. You ripped on Ubuntu for being steeped in politics, while ignoring the available facts, i.e. that Ubuntu 10.4 is a stable Long Term Support release, and that it wipes the floor with FreeBSD 8.0 in performance.

That's pretty funny. I mean, if you did it on purpose.

Re:Benchmarks...

[AusIV]

I am curious how much difference there would have been on a database system that would actually be running in a server environment. Ubuntu definitely got owned by that metric, but SQLite is hardly exemplary of a production database.

Re:Benchmarks...

[rubycodez]

you're sounding like someone who has one type of application in mind. my employer mainly deals in systems that handle money, speed is NOT the primary consideration, nor is memory or cpu foot print as hardware is such a small part of project budget they will get however much many boxes and however much cpu/ram the application wants. Anything in the distro that supports the app will be installed from packages, no compiling of distro-provided things will be done except in the most extraordinary situations.

Priorities are certfied/supported by application and driver/device vendors, long term support time, stability, ease of management. This is typical financial business mindset.

Re:Benchmarks...

[dkleinsc]

the big bad FSF

I must be missing the part where the FSF became the evil empire who wanted to control the world of computing, rather than a rebel alliance who wanted the users in control of the code running on their own computers. And the part where the FSF did anything to actively stop FreeBSD from doing its thing.

The Ubuntu project, the Linux project, and the FreeBSD project all have the same basic goal: create freely shareable really awesome (fast, effective, efficient, robust, useful, and as bug-free as possible) code that solves problems for people. Can't we all just get along?

Re:Benchmarks...

[nimbius]

Hey! benchmarks are too useful!

without these numerically-backed conslusory points, id be forced to attempt to justify budgeting and purchases to the PHB using the potential usefulness and benefit of a service to its userbase. giving an old man a line-graph and pie chart with made up numbers means he doesnt need to know anything more than red is bad, green is good.

Re:Benchmarks...

[Galactic Dominator]

A bit dated,

http://people.freebsd.org/~kris/scaling/os-mysql.png

Re:Benchmarks...

[zippthorne]

Ubuntu has a separate 'server' version (which really just includes a different set of packages and a different kernel build.)

Wait.. what else is there?

Re:Benchmarks...

[tarius8105]

How often is this important? I can think only of a few situations, such as when fitting a system into a small/cheap flash.

Its pretty important. When you are talking enterprise servers, I really dont want X running, or the thousand other things Linux has installed. Bare metal does not mean small like putting on a flash drive. I can create a bare metal server through the install options and then add in what is needed as oppose to having to strip it down post.

Re:Benchmarks...

[Murdoch5]

I dont actually just have one type of application in mind, I do lot of different work on my computers. No matter what I'm running I wouldn't go for anything but speed / foot print. When I built / designed my final year project in College my software was built for Speed!, there's literally nothing else that can get as important.

Re:Benchmarks...

[everynerd]

Ubuntu is a desktop operating system. Joe Blow doesn't care about anything you've posted. And if you're considering Ubuntu for anything but a desktop computer, then you're doing it wrong.

Re:Benchmarks...

[Just Some Guy]

When I built / designed my final year project in College my software was built for Speed!, there's literally nothing else that can get as important.

I bet your coworkers just love seeing a big commit with your name on it coming down from version control.

Re:Benchmarks...

[Abreu]

Ubuntu Server Edition does not install X by default

Re:Benchmarks...

[Abreu]

I must be missing the part where the FSF became the evil empire who wanted to control the world of computing, rather than a rebel alliance who wanted the users in control of the code running on their own computers.

The change probably happened when Obama became president and suddenly transformed from a "beacon of hope and change" to "a scary communist who wants to take away your freedom"

Re:Benchmarks...

[Abreu]

Canonical seems to disagree

Re:Benchmarks...

[Murdoch5]

Writing fast / optimized code doesn't mean writing hard to read code, it just means writing the best code.

Re:Benchmarks...

[Just Some Guy]

Writing fast / optimized code doesn't mean writing hard to read code, it just means writing the best code.

That's just darling! For some reason, I picture you writing an MD5 library that keeps a hash table of inputs to their computed outputs so that you won't have to recalculate them. That'd be fast! And optimized!

Re:Benchmarks...

[gmack]

Ext4 is hardly exemplary of a production file system and has known performance bottlenecks that seem to never get fixed. What it does have is good backwards compatibility with ext2 and ext3. ReiserFS, XFS, BFS would all have done better on that benchmark.

Re:Benchmarks...

[TheVelvetFlamebait]

You sounded like you had a good point, until you let fanboy rants cloud it.

Hey, politics ain't just for the mainstream, y'know.

Re:Benchmarks...

[Cato]

You can also install Ubuntu from a mini ISO (from Ubuntu.com) which gives you a really minimal CLI system, from which you can build things up. I did this on a home server recently and it ran comfortably in 128 MB.

Re:Benchmarks...

[BitZtream]

How often is this important? I can think only of a few situations, such as when fitting a system into a small/cheap flash.

Its important to those of us that know by not including extra stuff on the machine, that extra stuff can not possibly be exploited, cause problems, waste resources or any number of other reasons that don't come to mind right at the moment. Including extra stuff is almost always a bad idea when you're doing a server. Desktops are a little different. Interestingly enough, Ubuntu is desktop focused with some server on the side. FreeBSD is server only, with some people who refuse to give up on making it a desktop OS, which they actually do a reasonably good job at if you're technical enough to deal with some of the quirks along the way.

You are of the same mentality as MS. Disk space and memory is cheap, who cares if we waste space with unneeded stuff!

Thats fine for a desktop OS where making things work off the bat for the user is more important than being intelligent. Personally, I'd rather you give me a ton of options and sane defaults rather than an installer that basically just copies everything from the cd/dvd/nfs mount to the drive and reboots.

However, its certainly doable with Ubuntu, it just takes more work than with FBSD. The difference is really in the installer. Ubuntu is far more ignorant-user-friendly than FreeBSD, which is perfectly acceptable as its aimed more at being easier to use. I've been using FBSD since 2.2.x and with the exception of a minor change allowing for selection of your X desktop environment the FBSD installer hasn't changed since then and I still made a mistake and wasted an install during an install of an 8 preview. Entirely my fault for not paying attention, but its one of those things that the Ubuntu installer most certainly would have red flagged or just not allowed at all. Different designs for different targets, both are better than the other in the areas they intend to be.

Re:Benchmarks...

[ducomputergeek]

But in my experience I've never really seen many people use Ubuntu in the server capacity, they tend to use Debain if they are deploying on a server. Still, I love the fact that there is DTrace and ZFS in FreeBSD. Hopefully with some improvements in FreeBSD 8. We're still running FreeBSD 6.3 around here on our production machines. But improved ZFS support would make a case to upgrade.

Re:Benchmarks...

[nxtw]

Its important to those of us that know by not including extra stuff on the machine, that extra stuff can not possibly be exploited, cause problems, waste resources or any number of other reasons that don't come to mind right at the moment. Including extra stuff is almost always a bad idea when you're doing a server.

Just how much extra software is included in a base server installation of RHEL or Ubuntu?

You are of the same mentality as MS. Disk space and memory is cheap, who cares if we waste space with unneeded stuff!

It is not a Microsoft mentality. It's simple economics.

Personally, I'd rather you give me a ton of options and sane defaults rather than an installer that basically just copies everything from the cd/dvd/nfs mount to the drive and reboots.

Did you ever consider looking for a server installation image?

However, its certainly doable with Ubuntu, it just takes more work than with FBSD. The difference is really in the installer. Ubuntu is far more ignorant-user-friendly than FreeBSD

It doesn't seem like you've used Ubuntu besides of the desktop live CD. Ubuntu has separate server installation images with the Debian installer; it does not install desktop components. The desktop live CD basically copies the contents of the CD to disk.

I've been using FBSD since 2.2.x and with the exception of a minor change allowing for selection of your X desktop environment the FBSD installer hasn't changed since then and I still made a mistake and wasted an install during an install of an 8 preview. Entirely my fault for not paying attention, but its one of those things that the Ubuntu installer most certainly would have red flagged or just not allowed at all.

The Ubuntu server installer does not force you to install X at all.

Re:Benchmarks...

[value_added]

2. Varies, FreeBSD doesn't have perfect docs either and other distros do have better docs

Sorry, but FreeBSD's documentation is without par. Perhaps you're confusing the combination of info pages (mostly unwritten), non-existent manpages, Google searches, user forums and third-party wikis with "perfect documentation"? To the extent that "perfect documentation" actually means anything.

4. Yes, it does, there's a 10MB barebones installer for the dedicated and if you need less, pick another distro, even FreeBSD will probably be tons of tweaking at this point.

An installer? Tons of tweaking? LOL. You really don't know what you're talking about.

Re:Benchmarks...

[agnosticnixie]

2. Ah, yes, these nonexistent manpages I spent every other day browsing when i first started using linux. I must have imagined them. Is ubuntu's documentation less good? sure - but again, I question the FreeBSD manpages' reputation as "without par"

4. Do tell, please, why then the market is divided between BSD variants and Linux in the embedded world, could it be that, gasp, it actually scales down. Of course this is about Ubuntu, which is not build for ricers but for desktop use mostly.

Re:Benchmarks...

[VGPowerlord]

He meant to say a different default set of packages. They still use the same package repository, so all the packages are available to both editions.

Re:Benchmarks...

[selven]

That kind of table works both ways. Prepare for a nice friendly visit from the NSA.

Re:Benchmarks...

[micheas]

Do you know if that was pre or post "google patches" being donated to mysql?

Re:Benchmarks...

[micheas]

Last I heard Etrade was still running gentoo. (The etrade website is emerged and the whole system is run through qa and then pushed live.)

When I was asked to make computer purchases for a brokerage firm none of your issues were what I used to decide on a choice. It was ROI, Risk of failure over the life time of the hardware, and competitive advantage, if any. (I ran the numbers six ways till sunday and swallowed my morals and said by microsoft for the trading stations, (actual loss for a crash in a fast moving market vs, the 3x cost of not using Microsoft,) Everything is about money in Banks and Brokerage Houses, I don't think ease of management would have made any impression on anyone unless you could explain how it improved their Christmas bonus.

Re:Benchmarks...

[Galactic Dominator]

I'm not really sure, but I do remember this was right around the time 7.0 was released, and was shortly after the introduction of the completely fair scheduler.

Re:Benchmarks...

[tarius8105]

There were tests a few years back they did "how low can you go" testing. They were able to run freebsd 6 on a P200 with 32mb of ram.

I even remember running FreeBSD 2 on a P60 with 16mb of ram and only 400mb of disk, and it ran like a champ.

Ubuntu *is* Linux for a lot of people

[elrous0]

When my friends ask me about Linux, I usually steer them toward Ubuntu first, as it's the most user friendly and well-supported distro out there. Canonical really puts a lot of development work into it, and it shows (in this result and many others). In the past, I usually avoided the Linux topic altogether, as there were so many confusing distros that even trying to explain the concept of Linux to non-geeks (and even many geeks) was a huge pain in the ass. So, I for one welcome our new Ubuntu overlords.

Re:Ubuntu *is* Linux for a lot of people

[psm321]

And every time I try Ubuntu "one more time", I am amazed that people call it the most user-friendly distro. I can never get anything to work in it. My go-to "it just works" distro right now is Fedora, and that's what I recommend to people because things actually work. Of course most of the time I personally run Gentoo or Slackware, but there are occasions when I want things to just work.

Re:Ubuntu *is* Linux for a lot of people

[bostei2008]

My girlfriend just bought a new laptop and, not wanting to spend money on Windows Vista, installed Ubuntu on it.

She is a pretty normal user (what you would call a slightly above average Windows user), but she was amazed at Ubuntu. She got almost everything to work, and those things which would not work out of the box fired her up to get it to work.

Mostly she liked the all-around-friendly atmosphere of Ubuntu, the decidedly non-elitistic image, the helpful community. Also the fact that Ubuntu is pretty gorgeous helped.

For me it seems that the few problems she fought with had to do with missing drivers and video codecs. Also, for a new user, the package system takes some getting used to.

Re:Ubuntu *is* Linux for a lot of people

[psm321]

I think perhaps the problem is that techie users expect different things even out of a "user-friendly" distro than average users. For example, my dad figured out how to install a font on the eeepc default distro, while I was still completely stumped.

Re:Ubuntu *is* Linux for a lot of people

[Bill Dimm]

Wow, so I'm not the only one that doesn't understand the Ubuntu love-fest? I only tried it once (8.04 64-bit), but I got frustrated with it very quickly. For example, I logged in as a normal user (not root), selected the network configuration app from the menu, and was not prompted for the root password. Everything just came up ghosted and unusable. I tried to log in as root, but you can't do that ("admin not allowed to log in from this screen" -- is there some other screen that admin can log in from?). I ended up having to pull up a shell, guess the name of the network admin app (/usr/bin/network-admin), then su and run it. This is supposed to be user-friendly? How does something that brain-damaged get released? I ran into several other problems (it's been too long for me to remember details), and just gave up on it after a few hours. I haven't had problems like that with other distros (OpenSUSE has worked quite well for me lately). I just don't see why people think Ubuntu is so much better than everything else.

Re:Ubuntu *is* Linux for a lot of people

[Abstrackt]

Of all the distros my wife tried Ubuntu is the only one that didn't frustrate her enough to reboot the computer into Windows. Don't get me wrong, there are a lot of great distros out there, but Ubuntu is the first one my wife asked me to install on her laptop.

Re:Ubuntu *is* Linux for a lot of people

[agnosticnixie]

The problem is that user-friendly is measured using average users as a yardstick, with the assumption that more user-friendly = less techie user can get by. Which leads to the problem that the relatively average-user-friendly Ubuntu is pretty easy to hose if you stray away. For all it's worth, imo, Geek-proofing an OS is pretty much impossible.

Re:Ubuntu *is* Linux for a lot of people

[psm321]

I understand that there's perhaps some such effect there that I'm missing. But when even simple things like installing a package or trying to configure the network don't work, I find it hard to imagine how my expectations in that case are any different from an average user's.

Re:Ubuntu *is* Linux for a lot of people

[Bill_the_Engineer]

Wow, so I'm not the only one that doesn't understand the Ubuntu love-fest?

This should be interesting... Someone who doesn't understand why everyone doesn't think alike.

I only tried it once (8.04 64-bit), but I got frustrated with it very quickly.

So we all know how qualified you are when it comes to unbuntu. I use 8.04 since it's the latest LTS version and I haven't had any problems with it. Your mileage may vary.

For example, I logged in as a normal user (not root), selected the network configuration app from the menu, and was not prompted for the root password. Everything just came up ghosted and unusable.

Well maybe it would have helped to press that button label "Unlock" and then type your password.

I tried to log in as root, but you can't do that ("admin not allowed to log in from this screen" -- is there some other screen that admin can log in from?). I ended up having to pull up a shell, guess the name of the network admin app (/usr/bin/network-admin), then su and run it. This is supposed to be user-friendly? How does something that brain-damaged get released?

Again it would have gone a whole lot easier if you just clicked the unlock button. Quit being a drama queen.

I ran into several other problems (it's been too long for me to remember details), and just gave up on it after a few hours. I haven't had problems like that with other distros (OpenSUSE has worked quite well for me lately).

Whew... I was on pins and needles wondering if you would find a Linux distribution that you liked. *sarcasm*

I just don't see why people think Ubuntu is so much better than everything else.

Maybe because people have different tastes? A much better question you should have asked is "If everyone else thinks Ubuntu satisfies their needs, what am I doing wrong?" or the best question to ask is "I wonder if I should ask for help since I seem to be in the minority of people who think Ubuntu is too difficult to use?"

I heard some valid criticisms about Ubuntu, but not being user friendly isn't one of them.

I thought people who get caught up in the Windows vs. Linux vs. OS X vs. BSD were pretty small minded. How small a mind do you have to have to fight over which distribution of a single OS is the best? Heaven forbid someone else uses something different then you.

My apologies for making these snide remarks, I'll be in a much better mood once my coffee kicks in. Anyway, what kind of response did you expect?

Maybe you should have reworded your comment into:

I wonder if they will post a benchmark comparing FreeBSD 8.0 to my favorite linux distro?

Re:Ubuntu *is* Linux for a lot of people

[Abreu]

Mostly she liked the all-around-friendly atmosphere of Ubuntu, the decidedly non-elitistic image, the helpful community.

This, rather than any technical consideration, is what makes Ubuntu successful

Re:Ubuntu *is* Linux for a lot of people

[witherstaff]

I recently wanted linux on a HP laptop. I tried the latest Ubuntu disk and it just didn't load. I decided to try Mandriva and things worked great. It was the first linux install that I didn't have to do any tweaking for video or wireless. I was pleasantly surprised.

Re:Ubuntu *is* Linux for a lot of people

[pinkj]

I have tried Ubuntu three times. And each time I've had problems with video and/or audio and updates.

I have an older ATI AGP x850 video card which ATI's Linux driver doesn't support and the open driver doesn't cut it for gaming at all. This isn't the Ubuntu team's fault, but it doesn't help with wanting to try any Linux distro really.

I also have an M-Audio Fasttrack Pro audio device that caused problems in Ubuntu, but the latest release had fixed those.

And the major updates always break something. It's usually permissions problems for some reason. Maybe I'm doing something wrong, but I don't tinker in the CLI enough to cause havoc. I would usually tinker with my video related confs to get the damn thing working.

So I've given up, I still have my WinXP SP3 installed which simply works most of the time. It's not ideal, but at least I can use Pro Tools and play games with pretty graphics once in a while.

Re:Ubuntu *is* Linux for a lot of people

[Bill Dimm]

So we all know how qualified you are when it comes to unbuntu.

This thread is about whether or not Ubuntu is user-friendly and "just works." If you need extensive experience with it in order to be able to use it, it has already failed the user-friendly test.

Well maybe it would have helped to press that button label "Unlock" and then type your password.

Please keep in mind that this discussion is about it being user-friendly. I've seen lack-of-privilege problems handled in other distros two ways:
1) Tell the user he/she doesn't have the privileges required to do what he/she is trying to do.
2) Prompt the user for the root password.
Both approaches make it obvious to the user what the problem is, and what they need to do to proceed. Is it really supposed to be obvious that everything is ghosted in the application because it needs to be "unlocked"? If I click the "Help" button, do you know what it tells me about the "Unlock" button? Absolutely nothing. In fact, it says "When you start the Network Administration Tool, you will be prompted for the administrator password..." which is most definitely not what happens. To make you happy, I just clicked the magical "Unlock" button. Do you know what happened? The window hung for 25 seconds (would not respond to mouse clicks), then a dialog popped up saying "Could not authenticate. An unexpected error has occurred." I was never prompted for the root password. Very helpful. If you honestly cannot acknowledge that this is a disaster from a usability standpoint, there is really no point in even talking to you, because you are lost in fanboy land.

How small a mind do you have to have to fight over which distribution of a single OS is the best? Heaven forbid someone else uses something different then you.

Take a look back at the beginning of this thread. Hell, just look at the subject line: "Ubuntu *is* Linux for a lot of people." I'm not at all arguing that everyone should use what I'm using. I'm arguing against the idea that everyone should automatically use Ubuntu. Specifically, I'm arguing against Ubuntu being the one and only user-friendly distro. Other distros were very user-friendly and "just worked" before Ubuntu even existed. I can hardly remember the last time I had to vi XF86Config. I'm not saying people shouldn't use Ubuntu if they are happy with it, I just don't see why Ubuntu gets so much credit for being user-friendly while the other distros get so little, and your post didn't add a single thing to that discussion.

Re:Ubuntu *is* Linux for a lot of people

[Bill Dimm]

and was not prompted for the root password.

you can do this on any distro, its a question of changing the privilege flags on the applications file (probably in /usr/bin)

I think you've misunderstood what I was saying. I wasn't saying that the app was suid root. It wasn't. The app needed root privilege if you wanted to change the settings, and I didn't have root privilege. Most other distros will prompt you for the root password in a case like that, but Ubuntu didn't, it just failed to work. If you click the "Help" button, it says that it will prompt you for the root password, but it doesn't. See more detail here.

did you even bother to look at the community documentation or the forums? if you run in to a problem 200 other people have too and more often than not someone smarter than you finds a way to fix it

No, I didn't turn to the community (and the built-in documentation was wrong). I wasn't hell-bent on running Ubuntu; I just wanted to see what all of the fuss was about. I'm sure it can be made to work with enough effort. The point of my post is that I don't understand why Ubuntu has a reputation for being so much more user-friendly than all of the other Linux distros. If you have to turn to the community and spend a lot of time mucking around to get basic functionality to work, it really isn't user-friendly. Maybe my experience was a fluke, but I've found the other distros to be much less of a hassle.

Re:Ubuntu *is* Linux for a lot of people

[FrankieBaby1986]

FAIL.

is there some other screen that admin can log in from?

The whole idea is that users DO NOT LOGIN AS ADMIN, EVER. In fact, with ubuntu, you should not even have an admin password at all.

sudo provides the functionality you need, as you discovered.

Now as to why your network configs were locked, I haven't a clue. Maybe there was a bug in 8.04 (I was only on it a short time, but don't remember that issue) Or perhaps something went wrong with your install. But I do seem to remember there being an "unlock" button to press occasionally.

In 8.10, network manager handles networking configuration, and network-admin only deals with proxy settings, so I'm not sure what you are talking about.

But really, our anecdotes are kinda useless, and perhaps some of the issues you had were growing pains from 64 bit (lots of things still have trouble, esp on Windows.) Just speculating, as you didn't remember specific problems you had.

Re:Ubuntu *is* Linux for a lot of people

[wm_brant]

Well, Ubuntu worked for me. I recently sent my last Windows PC off with my daughter to college, and built my own PC. Taking a leap of faith, the new PC got 64-bit Ubuntu from the start -- no Windows -- and everything is, and has been working since I first powered it up. This, with no real *nix experience.

I understand that Ubuntu might be considered a Linux for neophytes, but that's OK with me. I needed to be successful with *nix since I was starting with bare iron and zilch experience, and Ubuntu delivered that. If at some point I may choose to go with another distro, and I know that much of what I learn with Ubuntu will apply to any *nix distro. But for now, I'm content.

Have I had issues? Yup. But there is a lot of help for Ubuntu around. There might be a lot of help for the other distros, too, but I knew there was a lot of help for new Ubuntu users before I even started. I knew a lot of people have been successful with Ubuntu. I knew that my hardware was compatible. Basically, I had a -- maybe unjustified, and maybe myopic -- warm and fuzzy feeling about Ubuntu.

-- Bill

Re:Ubuntu *is* Linux for a lot of people

[Bill_the_Engineer]

Take a look back at the beginning of this thread. Hell, just look at the subject line: "Ubuntu *is* Linux for a lot of people." I'm not at all arguing that everyone should use what I'm using. I'm arguing against the idea that everyone should automatically use Ubuntu. Specifically, I'm arguing against Ubuntu being the one and only user-friendly distro.

While the person creating the thread may have been a little over zealous about his feelings for Ubuntu in the title. I didn't take it as an insult to the other distributions. I took it more as something he is more familiar with and more comfortable in recommending. He did say:

I usually steer them toward Ubuntu first, as it's the most user friendly and well-supported distro out there.

That means that he may have recommended other linux distributions...

It was Canonical that decided early on into making user friendliness as its top priority for Ubuntu while keeping the distribution free to download and distribute. It was this attitude early on that gave Ubuntu the perceived lead in the desktop area.

I can hardly remember the last time I had to vi XF86Config.

I have a dual monitor set up running proprietary Nvidia drivers and the latest incarnation of Ubuntu works without forcing the user to manually edit the XF86Config.

When 8.04 came out, I did have to edit the XF86Config because the nvidia driver was looking for an option. But truth be told, I had to edit the XF86Config with the other distros too.

I'm not saying people shouldn't use Ubuntu if they are happy with it, I just don't see why Ubuntu gets so much credit for being user-friendly while the other distros get so little,

Mostly because Canonical invests in spreading the word about Ubuntu, and Ubuntu doesn't have a version that requires a paid subscription. Some people feel (possibly wrongly) that a particular distribution sold out to Microsoft at the expense of the Linux community. Some distributions exists solely to offer a free version of RedHat enterprise. While other distros requires the user to perform a distro-upgrade every six months. There are a multiple reasons.

As with all things, if there are multiple choices being offered each person is free to chose and there will be a perceived winner and losers. Fortunately for the distros, each distro was built to satisfy a need for a particular subset of the Linux community and profit wasn't the driving force behind most of them.

However given enough time, the number of distributions will fall and the market will decide which will be the mainstream choices.

your post didn't add a single thing to that discussion.

Right...

Every distribution has its edge case, and you found a workable solution in something besides Ubuntu. I was chastising you for trashing someone else's choice. You will also need to tolerate the fact that your particular distribution isn't the popular flavor of the month.

Re:Ubuntu *is* Linux for a lot of people

[Bill Dimm]

FAIL.

is there some other screen that admin can log in from?

The whole idea is that users DO NOT LOGIN AS ADMIN, EVER

You missed the point of the parenthetical comment that you quoted. By Ubuntu giving the error message "admin not allowed to log in from this screen," it implies that there may be some other screen where it is OK to log in as root. If it is never possible to log in as root, then that is what the error message should say, instead of saying something confusing about "this screen." The point of my post was to question whether or not Ubuntu is really so much more user-friendly than other distros, and having confusing error messages isn't user-friendly.

But I do seem to remember there being an "unlock" button to press occasionally

Please see my post here where I talk about the "Unlock" button.

perhaps some of the issues you had were growing pains from 64 bit

Congratulations on being the first person in this thread to propose a legitimate potential explanation for my experience with Ubuntu being so different from the apparent norm, instead of choosing to attack me personally. I had actually considered that as a possibility all along. That is why I made a point of specifying that I had used the 64-bit version, so if the 64-bit version was known to be crap, people could point that out as the explanation. On the other hand, I was already using 64-bit OpenSUSE before I ever touched Ubuntu, and had no problems with it at all. So, it is certainly possible to put out a working 64-bit version, and has been for years. But, if the majority of Ubuntu users are on 32-bit, and the 32-bit version works great while the 64-bit version is broken, that may very well explain why so many people love Ubuntu while my experience with it was such a disappointment.

Re:Ubuntu *is* Linux for a lot of people

[Bill Dimm]

Doesn't work. See this post for details.

Re:Ubuntu *is* Linux for a lot of people

[psm321]

So apparently the Ubuntu fanboys don't like me mentioning that their distro doesn't actually work... my first flamebait mod ever :)

Awful

[Anonymous Coward]

FreeBSD rather ends up taking a wallop to Ubuntu Linux, but there are a few areas where FreeBSD 8 ran well.

lrn2preposition

For those that want to skim TFA for the bad result

[MikeRT]

The areas where FreeBSD gets its ass kicked by Ubuntu start on page 7...

It seems to me like FreeBSD's real problem is incredibly bad I/O compared to Linux. The majority of the CPU-heavy tests were nearly neck-in-neck.

Safe to assume?

[Capsy]

Maybe this is a sign that more systems will start coming with Ubuntu already equipped?

Re:Safe to assume?

[Ian Alexander]

Yes, because the masses of casual users struggling under the crushing weight of FreeBSD that was preinstalled on their PC will naturally flock to Ubuntu...

Re:Safe to assume?

[selven]

And although Windows users keep parroting supposed advantages in software compatibility they simply fail to see what customers really need - actual security, ability to function without being an admin, and a built-in package repository and management system (Windows geeks simply can't grasp the fact that your average Joe does NOT enjoy chasing dependencies and would prefer the system to do things for him), 2009 will not be the year of Windows on the desktop.

Re:Sigh...

[Timothy Brownawell]

My Karma is going to take a hit for this

What karma? You're already posting at +0, even with being a subscriber...

When there's something you need to do that can't be done with Windows but can be done with Lunix, keep in mind that you can do an even better job with Mac OS X.

Except true multitasking? That shared menubar seems to assume you're only doing one thing at a time, and Windows' lack of proper focus-follows-mouse (ie, without raise-on-click) likewise.

Some argue that BSD can do it better but no one makes software for BSD since no one gives a flying fuck.

I was under the impression that the BSDs could generally run Linux binaries (some sort of compatibility mode thing)?

With all seriousness in mind, BSD isn't useful for anything really.

What about as an embedded OS for consumer electronics?

Wallop to?

[jepace]

Is taking a "wallop to" something a good thing or a bad thing? I guess that's one way to make sure we read the story is to have the headline make no sense...

Re:Wallop to?

[BitZtream]

Not American? change 'wallop' to 'hit' and you'll get the point I think. Never really thought about it, but I'd say 'wallop' is probably old slang from something else.

Re:Wallop to?

[ZERO1ZERO]

what about codswallop? Or is that something else entirely?

Truly crap-tastic charts

[vrmlguy]

Eight pages of bar charts, each gray-on-gray. On half of them, shorter bars mean better performance, on the other half, longer is better; the only way to know which is which is in a legend, written in a small font.

Here's a suggestion: color-code the bars! Green is good, red is bad, yellow is in the middle of the road. For bonus points, choose the saturation based on magnitude of the differences. If the numbers are close, go with grayer bars, if the differences are dramatic, use dramatic colors.

Finally, how about a line chart at the end showing all of the numbers in one place? Yeah, you'd need to convert everything to be consistent if longer or shorter is better, but that's a good idea anyway.

Re:Truly crap-tastic charts

[randomsearch]

I agree the charts are poor. What about the fact that the axes don't necessarily start at 0? This can be misleading, in that a small difference can be made to look much more significant. Broadsheets in the UK do this all the time.

One point... red-green colour blindness is quite common, so if we're being pedantic you'd probably want to avoid using them for contrast. Problem with colour-coding is also deciding what is "good" and "bad", which is making a subjective judgement.

RS

Re:Truly crap-tastic charts

[dkleinsc]

What I'd recommend for displaying this data is switching from charts to tables, and use alternating light green and white rows (little holes on either end for dot-matrix printer alignment optional).

Re:Truly crap-tastic charts

[vrmlguy]

I agree the charts are poor. What about the fact that the axes don't necessarily start at 0? This can be misleading, in that a small difference can be made to look much more significant. Broadsheets in the UK do this all the time.

That's one of the first tricks I learned from How to Lie with Statistics. The other is that log scales should be used much more than they actually are. (Google finanace gets this right, although it's not always obvious.)

One point... red-green colour blindness is quite common, so if we're being pedantic you'd probably want to avoid using them for contrast. Problem with colour-coding is also deciding what is "good" and "bad", which is making a subjective judgement.

Have you ever seen the Map of the Market? It's a Java applet that gives you the choice of red-vs-green (the traditional subjective value judgment, at least in the West) or blue-vs-yellow (especially useful if you have red-green colorblindness, its high contrast is probably useful for lots of other forms as well).

BSD did rather well

[John Jamieson]

I'm not a BSD user, but I don't see BSD taking a real kicking in these benchmarks. In the majority of the benchmarks, the average user could not discern a speed difference.

Re:BSD did rather well

[Anonymous Coward]

OS benchmarks are misleading, usually. Claims such as "12% decrease in performance" (of linux kernel..) usually translates that kernel execution time takes ~ that much longer, but neglects that majority of time kernel is idle and userspace process (ab)uses the processor (and whatever else). So, 12% is probably 1% in real usage case. I guess same applies to sensational Windows 7 speed improvement over "slow" Vista.

Re:BSD did rather well

[canajin56]

Look, this is benchmarking. You need to know your lingo. Take 3D benchmarking. Even though most monitors will have your FPS capped at 60 Hz, if one card gives you 250 FPS and the other only gave you 240 FPS, you say that the new card "absolutely destroys the competition," for example. Anything even partially outside the error bars translates into at least a "solid thrashing."

Re:BSD did rather well

[ichthyoboy]

Anything even partially outside the error bars translates into at least a "solid thrashing."

When did Phoronix start using error bars? I've never seen a single one on any of their graphs, and it makes me really suspect of their conclusions on performance differences.

Re:BSD did rather well

[viridari]

The machine that they performed this test on was far from an average case. A difference of two seconds between FreeBSD & Ubuntu on this box might equate to a 10+ second difference on a typical Core 2 Duo desktop, depending on what kind of test we're talking about.

Benchmarks on superlative hardware don't impress me. Have a budget of $500 for a computer, go to Wal-mart and buy it, and run the benchmark on that. That's going to be far more typical of what most people have on their desks.

Re:For those that want to skim TFA for the bad res

[amorsen]

It's fairly difficult to bungle traditional CPU-heavy loads. The kernel just needs to get out of the way and let userspace do whatever it wants to do. Try the same tests on Mac OS X and Windows (assuming they compile), and you'll see just about the same performance.

Re:Sigh...

[hedwards]

You'll have to explain that to Codeweavers, the Win4BSD people and the other commercial vendors that make software for it. Sure there probably isn't as much stuff aimed at home users as for Linux, Mac or Win, but there is a substantial amount of commercial activity involved. It's odd that some random company would fund a new routing architecture if it's not useful for anything.

Re:Sigh...

[mikeypimp]

Wait, you actually use focus-follows-mouse? You're a better man than I.

performance over the years

[Danathar]

I'd actually like to see a chart with these types of performances going back a number of years. It would be interesting to see the difference between RH10 and the Latest Fedora release (assuming you could even get RH 10 to run on modern hardware without backporting drivers). Same wit FreeBSD.

Re:For those that want to skim TFA for the bad res

[Galactic Dominator]

That's to be expected considering the defaults of ext4 vs ufs2. You can increase flush time on ufs2 and expect a similar increase. Revert to ext3 and it would be a completely different outcome. Interesting to see all the chest pounding on choice for default settings in a desktop enviro vs a traditionally server one. Would have been a been comparsion to use the upcoming PCBSD's release vs Ubuntu's, but we've seen the bias from Phoronix before.

FreeBSD is still in debugging mode

[Just Some Guy]

From the update notes in /usr/src/UPDATING:

NOTE TO PEOPLE WHO THINK THAT FreeBSD 8.x IS SLOW:
FreeBSD 8.x has many debugging features turned on, in both the kernel and userland. These features attempt to detect incorrect use of system primitives, and encourage loud failure through extra sanity checking and fail stop semantics. They also substantially impact system performance. If you want to do performance measurement, benchmarking, and optimization, you'll want to turn them off. This includes various WITNESS- related kernel options, INVARIANTS, malloc debugging flags in userland, and various verbose features in the kernel. Many developers choose to disable these features on build machines to maximize performance. (To disable malloc debugging, run ln -s aj /etc/malloc.conf.)

Since the article didn't mention anything about disabling all the debugging options, I'll consider this an invalid benchmark until shown otherwise.

Re:FreeBSD is still in debugging mode

[Anonymous Coward]

WITNESS and INVARIANTS were disabled on Sep 10 (http://svn.freebsd.org/viewvc/base?view=revision&revision=197065). FreebSD 8.0-RC1 was built on Sep 17th.

Re:FreeBSD is still in debugging mode

[magus_melchior]

I'm also wondering whether the Linux distro had newer libraries and drivers (esp. drivers). I noticed that the GCC compile had the BSDs using GCC 4.2 while Ubuntu was using 4.4; it might not be significant, but it could have made all the difference there.

Though it does seem (given other comments) that Linux handles I/O faster-- I'm not sure how much turning off debugging symbols will do, as 7.2 was also left in Ubuntu's dust there. It would certainly be worth checking out for the FreeBSD crew.

Re:FreeBSD is still in debugging mode

[mythz]

Since the article didn't mention anything about disabling all the debugging options, I'll consider this an invalid benchmark until shown otherwise.

Every operating system can be further optimized. What's being tested here are the 'Default Settings' as proposed by the installer directly off the Installation Disk.
'The Defaults' also tend to have the characteristic of being the most widely used.

Re:FreeBSD is still in debugging mode

[Just Some Guy]

Are those options enabled out of the box as a standard new-to-the-platform sysadmin would install and configure a firewall, Apache Webserver, or fileserver? If so...

They are if a standard new-to-the-platform sysadmin installs pre-release testing versions.

Look, we aren't all fulltime sysadmins -- some of us work for small shops and wear many hats, and do the best we can.

So install the normal release versions that don't have these quirks. :-)

I'll admit, if there's "advanced configuration options" (recompiling your kernel counts as advanced--especially since in the BSDs I've seen that will cost you *all* user support)

So you've only seen OpenBSD? I've never heard of that being a problem on the FreeBSD or NetBSD mailing lists. Now, people will want to see your kernel configuration file if you've changed stuff and are having problems, but it's sort of expected that you'll want to build your own kernel at some point and no one thinks much of it.

Re:FreeBSD is still in debugging mode

[Just Some Guy]

Every operating system can be further optimized. What's being tested here are the 'Default Settings' as proposed by the installer directly off the Installation Disk.

Until recently, those default settings for FreeBSD slowed it way down in the interest of collecting extra debugging data. That's no longer the case as pointed out by a couple of ACs above (and I hadn't updated to the most recent prerelease version to notice yet).

Re:FreeBSD is still in debugging mode

[kc8apf]

Except when the defaults of a release candidate have a bunch of debug options turned on that aren't normally enabled for actual releases. This wasn't testing the 'Default Settings', it was testing the debugging settings.

Re:FreeBSD is still in debugging mode

[Cato]

Mod parent up even higher... in fact mod down the ones of equal score, this is probably the only comment that matters. It's unlikely Phoronix disabled all the debug features (would probably need a system recompile) so the benchmarks are most likely a waste of time.

Re:FreeBSD is still in debugging mode

[BitZtream]

Cool, now what about all the system libraries (which are more important in some cases than the kernel itself) that still have debugging turned on.

Debugging enabled in libc is a killer for EVERYTHING on the system, regardless of the kernel specific options.

Re:Sigh...

[marcosdumay]

You should try. It is great to be able to work at windows without raising them.

Hard to take this seriously

[Just Some Guy]

I which I'd seen this before I posted my last comment:

FreeBSD 8.0 vs. Ubuntu 9.10 BenchmarksFreeBSD performed awful in comparison to Ubuntu when doing random writes, where the latencies were extremely high and off the charts compared to Mark Shuttleworth's operating system.

How "awful" for FreeBSD and good for the OS that Mark wrote. My grammar and clarity aren't always perfect, but hey, I don't get paid to write this stuff. :-)

Re:Hard to take this seriously

[SteveFoerster]

but hey, I don't get paid to write this stuff. :-)

Oh, admit it -- sure you are, it's just that it's by an employer who doesn't realize that's what he's paying you to do. ;-)

Re:Hard to take this seriously

[Just Some Guy]

Dude, we don't talk about Fight Club!

Phoronix? Moronix more like.

[LizardKing]

Yet again a benchmark against a pre-release version of FreeBSD where the testers didn't even bother reading the documentation. Anyone actually familiar with the FreeBSD development and release process would know that a release candidate has a considerable amount of debugging options turned on. This is to help diagnose any problems as the last issues are shaken out of a release, but has an adverse impact on performance.

Re:Phoronix? Moronix more like.

[kill-1]

Yeah, everytime I read one of their articles "Moronix" comes to my mind. But as long as they are the only ones who regularly benchmark Linux (and sometimes BSD) systems they have a kind of monopoly.

Re:Phoronix? Moronix more like.

[Conley Index]

Anyone actually familiar with the FreeBSD development and release process would know that a release candidate has a considerable amount of debugging options turned on.

On Sep-10, most debugging was disabled: http://lists.freebsd.org/pipermail/svn-src-all/2009-September/013399.html

On Sep-17, there was the last commit before 8.0-RC1: http://lists.freebsd.org/pipermail/svn-src-all/2009-September/013645.html

Anyone familiar with the FreeBSD development and release process would know that there are no fixed rules rules when certain stuff happens and there are no sweeping changes like turning off debugging between a late RC and the actual release. (Other debugging stuff like kernel and module symbols are kept for the release.)

Re:Phoronix? Moronix more like.

[rivaldufus]

Phoronix loves doing this. Last time, they compared a beta version of FreeBSD to a release of Ubuntu. The debugging definitely hurts performance.
I complained about this once, but a bunch of teenage kids started saying, "Ubuntu R0X0RS!"
I'm not sure where people got the idea that benchmarks are the one and only metric for comparing OSes. When Ubuntu doesn't win a benchmark in one of their OS shootouts, they'll have some explanation why that doesn't mean a lot.
I think they're still reeling from that time around 1999 or 2000 when NT4 trounced a redhat tuned linux in web server performance. It seems to me that OS X is the real desktop competition for Linux, and that OS X is winning "the hearts and minds" of the users. I'm sure Ubuntu can beat OS X (wasn't there a phoronix shootout about this?) in most benchmarks, but that doesn't matter if desktop users generally prefer OS X.
I suspect it would be useful to have more articles about what is needed to improve in Linux on the desktop... but to be fair, phoronix is just a benchmark, so that would be out of their scope.

terrible review methods

[buddyglass]

I stopped reading when I realized they didn't even use the same version of GCC in their compilation comparison.

Re:terrible review methods

[RegularFry]

To be fair, if they're comparing what you get out of the box, that's not unreasonable. In that case they'd be treating the OS as a complete package, apps and all, which is sort of the point.

Of course, they could just be ignorant numpties. You got further than I did by RingTFA.

Re:terrible review methods

[buddyglass]

Actually compilation itself is a questionable benchmark due to all the variables you can't control. For instance, maybe the code contains pre-processor directives that result in entire modules not being compiled on one OS or the other? But if you're going to use it as an OS benchmark, you should at least use the same version of GCC.

The other dumb thing I often see are benchmarks using tasks that are single-threaded and almost entirely processor-bound. In other words, tests that are mostly useless for exposing differences between operating systems, most of which revolve around scheduling, memory management and I/O.

Re:Sigh...

[timmarhy]

freebsd is 100% binary compatible with linux. so even if you don't have the source or are too retarded to build from source yourself, the linux binary will run without issue or performance penalty.

i find it ironic that the gp talks about no one making software for BSD but cites Mac OSX as being his prefered option. what a fucktard....

Interesting comparrison but

[Chrisq]

Interesting comparison but the SQL test was a bit off testing a database installed on ext4, a journalling filesystem. Anyone wanting intensive database work would use ext3 for the DB partition.

Also, it would be interesting to see whether FreBSD still outperforms Linux in low-memory situations. This used to be an area where BSD had a clear advantage.

Re:Interesting comparrison but

[perbu]

Why on earth would you choose the only filesystem with a broken fsync() for your database partition?

Re:Interesting comparrison but

[BitZtream]

UFS2 also does journaling, although using soft updates which generally does make it faster than the original non-journaling variant.

Re:Interesting comparrison but

[Zephiris]

UFS2 doesn't use journaling. Soft updates is entirely different. UFS2 has the option of using gjournal on FreeBSD to add journaling, but it's not the default option.

Phoronix using ext4 vs UFS2 (instead of the more direct stable/age-based comparison against ext3) is a bit stacking the deck as far as anything that touches I/O. They also don't mention what options they use/changed if any, but could've further stacked things on ext4 via mount options, and other people cheerily mention Phoronix never using 0 for the basis of axis.

They might as well have used ZFS (production) vs. btrfs (alpha) in such comparisons. In both cases, you'd be talking about filesystems that have been production stable for years, vs. 'the new schitzo kid on the block', whether for better or worse.

Re:Interesting comparrison but

[Haeleth]

It's not even that. I have no problem using SQLite on ext4. It runs fine.

The real problem, I suspect, is that they're probably running those INSERTs individually. With SQLite it's very important to batch your INSERTs in transactions, because it does an fsync() every time it commits a transaction -- and that's what's so expensive on ext4.

Tugboats

[Murdoch5]

First off Free-BSD is a sourced based thanks to Ports so you have the option of adding in optimization and Ubuntu is nativity binary based, not to say you can't get source out of apt. How can you compare a binary and a source, that's like comparing Gentoo VS Ubuntu. Gentoo will win 100% of the time because we you make sure that only use the fastest of the CFLAGS where in Ubuntu you need to have everything built into the binary. I just don't see a fair link between these two operating systems.

Personal experience with BSD shows me just how messed up that system is. I'm surprised they could even install the thing to bench mark with it, and then waited 5 days for ports to grind away getting the packages set up.

Re:Tugboats

[BitZtream]

First off Free-BSD is a sourced based thanks to Ports

I don't know about you, but I stopped using ports a long time ago when pkg_add -r worked for pretty much everything. I only use ports for bleeding edge packages now days, when I want an update that isn't available in ports directly anyway, and requires me to update the ports build with a newer source package from the original authors. If I need to debug or test software I'm developing on, the ports may come into play so I can change some compile options. I don't think gcc has ever been run on any of my production FBSD servers.

Re:Tugboats

[Murdoch5]

I stand corrected!!!, I didn't know there was binary options available for free thanks for letting me know that. Still would it not be a better idea to compare the kernels

Re:Tugboats

[Elshar]

Are you seriously trying to say that as a geek you couldn't figure out how to get FreeBSD installed on a system? Or that other geeks attempting to benchmark various *nix distros somehow can't figure it out?

It's really not rocket science here. The installer for FreeBSD is more or less a text (curses) based equivalent of Ubuntu's installer. Just (off the top of my head, the specifics I'm sure have changed over time) "express -> "A" for use entire disk" -> standard -> "A" for auto defaults -> "All" for install all the base packages (personally I usually do with just the bare minimum) -> choose CD or FTP install -> let it do it's thing.

You don't get much more hand-holding than that without using a GUI-centric *nix distro like Ubuntu, FC or SuSe.

Re:Tugboats

[Murdoch5]

Personal experience with BSD shows me just how messed up that system is. I'm surprised they could even install the thing to bench mark with it, and then waited 5 days for ports to grind away getting the packages set up.

That's not what I said with that paragraph, what it means is that the installer is badly designed. Now to be fair the Ubuntu installer isn't great but any one can figure it out. You never partition a disk using auto, unless you don't understand how to partition a disk and then why aren't you using windows.

The best installer is Gentoo, just extract the base system into the root partition and then use the package manager to install the rest.

Conclusions ?

[godrik]

The conclusion of the benchmark seem pretty unclear to me. comparing a development versin with debugging compiled with different compilers on different operating systems lead to different result.

I could have guessed that. I believe a benchmark is useless without a good insight of why the performance are different and how to correct the problem.

Re:For those that want to skim TFA for the bad res

[Bralkein]

FTFA:

With Ubuntu 9.10 we were using the x86_64 server CD of the Alpha 6 build. With FreeBSD not shipping with a desktop environment by default, we used the Ubuntu server CD so that both could be tested just from the terminal in a similar environment.

So they are comparing FreeBSD to the Ubuntu server version, but not really for the right reasons.

Re:For those that want to skim TFA for the bad res

[Galactic Dominator]

I suppose that's a bit closer, but it still doesn't address the differences between fs defaults. ext4 is far more aggressive than ufs2 is by default. A closer comparison would have be to gjournal ufs and mount async which would have been a relatively close comparison to ext4, but that still would not be a fair io comparison as gjournal scales incredibly well but can double write time on single write. I am struggling to comprehend ext4 as a default filesystem in a server(or anywhere) however, considering it still has crash corruptions issues.

http://en.wikipedia.org/wiki/Ext4
http://onlamp.com/pub/a/bsd/2008/02/26/whats-new-in-freebsd-70.html?page=3

Application comparison rather than OS comparison

[Itkovian]

How is this really comparing OSses? From the article, I gather that the first benchmark, Imagemagick compilation, compares two GCC versions plus the OS that performs the I/O on behalf of GCC:

"Starting off with the timed ImageMagick compilation process, FreeBSD 8.0 was able to build this open-source application noticeably faster than on FreeBSD 7.2 (even though both are using GCC 4.2.1), but Ubuntu 9.10 with GCC 4.4.1 took much less time to build ImageMagick."

Duh.

Wait, who won?

[Qubit]

FreeBSD rather ends up taking a wallop to Ubuntu Linux, but there are a few areas where FreeBSD 8 ran well.

Apparently taking a wallop to != bringing the hurt.

Perhaps it would be clearer to say For the most part, Ubuntu beat FreeBSD, or perhaps FreeBSD got walloped by Ubuntu.

Re:benchmark fail!

[Anonymous Coward]

You, sir, are and idiot. You're literally saying that all OS benchmarks must involve operating systems with identical designs. This wasn't a micro-kernel benchmark for thread concurrency, process efficiency, but a whole operating system benchmark -- And however suprising it may sound, the filesystem can actually be a part of the operating system... WOW! Why don't you take your "O/I" intensive tasks and go buy a book on anger management "you fucking moron".

Re:Sigh...

[Timothy Brownawell]

Wait, you actually use focus-follows-mouse? You're a better man than I.

Heh. Focus-follows-mouse and virtual desktops are actually one of the main reasons I run Linux (Debian/XFCE) at home, followed somewhat closely by apt-get/aptitude and less closely by the great command-line support (which if I ran Windows I could always get with Cygwin/MinGW or these days maybe powershell).

Re:Sigh...

[jedidiah]

You can only do an equally good job with MacOS.

For this grant priveledge you get a meagre selection of hardware that will
either gouge you or leave you wanting. You also get an alien environment
with a number of annoying quirks and inferior package management to any
Unix. Package managment is a "Unix thing" and not just a Linux thing. This
is one area where MacOS demonstrates it's not really Unix.

Some of the proprietary tools you get with MacOS might be moderately more
useful but they will have quirks of their own, suffer from NIH syndrome
and may also suffer from addressing problems in a superficial manner.

BSD is at least a proper Unix.

Its no surprise...

[interval1066]

Some of the code in BSD family of code streams is so old its growing fungus. Still, when I set up a server I usually reach for OpenBSD, its proven rock-stable to me after years of trouble-free mail, firewall/NAT, and web server service. But Ubuntu's my desktop.

Re:Its no surprise...

[DaMattster]

I do much the same as you. The BSDs make excellent servers, firewalls, and routers. Although, I do use PC-BSD for my desktop. Comes down to personal preference. I use OpenBSD for my firewall, router, and vpn appliance. FreeBSD serves up my web pages, ftp, dns, and just about everything else.

Re:Its no surprise...

[petrus4]

But Ubuntu's my desktop.

I don't understand this. I tried using Ubuntu for six weeks before moving to FreeBSD. For me, it was a nightmare. Bloated, slow, horribly unstable. I've heard any number of horror stories about what the update process can do to a machine, as well.

That was eight months ago, and since installing FreeBSD I haven't looked back.

Which Ubuntu version are you using?

Compiling in FreeBSD

[phorm]

Actually, you'll probably do quite a bit of compiling in FreeBSD too. While you can pull binary packages, most people I know prefer to use a package manager like portsnap and compile source packages.

If you have a script to take care of compiling the common stuff, then just "make config" everything that need user input and let it fly. Granted, most servers I've used BSD on have been dual or quad core Xeon/AMD PC's with a fair bit of RAM, so it is overall a fairly quick process, though it was still not too bad on a few P4-era celerons.

The nice thing about FreeBSD source-compiled apps that I truly did love compared to BSD is the little tweaks you could do to avoid tons of crap dependencies. Debian used to be fairly "clean" as far as deps, but both Ubuntu and Deb are now getting quite ugly in that you get rather unwanted stuff in order to get the package you want. In most FreeBSD stuff, for example, I can check the "No X Server" box and happily compile my apps without any X support whatsoever. On Debian/Ubuntu I end up trying to install some CLI system monitoring tool or CUPS, whatever, and end up with a whackload of x.org stuff because it's tied to some font which is loosely tied to the actual package I'm trying to install. On Ubuntu desktops, trying to exorcize the demons of "Evolution" without having it remove other important stuff due to deps is near impossible. Sometimes there's a separate branch for a "cleaner" install, but often enough not.

Of course, BSD ain't perfect. Linux tends to have a lot of "new" stuff that BSD is a little more "conservative" in bringing into the mainline (iSCSI support for example was a fairly recent addition compared to 'nix), but overall the package system is powerful indeed.

Having not used Gentoo (yet) but knowing others who have, it seems like it might be somewhat similar in concept. The issues I've heard with are mostly in people getting to the up-and-running stage, but - similar to BSD - avoid annoying little conflicts or unwanted cruft is a lot easier than Debian/Ubuntu's precompiled binaries. Of course, I've also heard of much frustration in both if you have find halfway through a *long* compile that you missed something you should have flagged/included.

Re:Compiling in FreeBSD

[nickname_clash]

alias "aptitude=aptitude --without-recommends'

Re:Compiling in FreeBSD

[TeknoHog]

The similarities are not just incidental, Gentoo Portage was shamelessly copied from BSD Ports.

FreeBSD and i/o monitoring

[phorm]

Other than the binary-vs-compiled issues, part of my personal love/hate relationship with BSD has been related to IO tools and issues.

I find that FreeBSD's resource-usage monitoring to be quite a bit nicer than 'nix. One example is just pressing "m" in "top" to get a per-CPU list of usage rather than the 300% = roughly 3-of-4 CPU's in heavy use.

Stats on disk IO have always seemed to work for me, whereas in 'nix I've tended to need more specialized drivers compiled in the kernel. It could be that I just got lucky and always had supported hardware, but it has been much less of a trial.

Managing IO on certain devices in BSD has been much more of a pain though at times. The drivers of certain less-expensive RAID cards (often used in Dell servers) have been a true pain to optimize. They don't actually break, but all seem to have different buffer-depths requiring special config to prevent overruns which lead to poor performance.

Re:FreeBSD and i/o monitoring

[BitZtream]

Other than the binary-vs-compiled issues, part of my personal love/hate relationship with BSD has been related to IO tools and issues.

What? I have about 10 BSD boxes under my control, I'd guess about half of them have never had a compiler run on them, and the other half probably have never installed a precompiled binary. Some of these machines are for development purposes where I want to tweak things (like enabling debugging in libraries and such) so I compile everything on them myself with my specific settings. The other half are internal servers, of which there is almost never a reason to use anything other than precompiled packages.

'pkg_add -r zsh' works pretty much the same as 'apt-get install zsh'

Managing IO on certain devices in BSD has been much more of a pain though at times. The drivers of certain less-expensive RAID cards (often used in Dell servers) have been a true pain to optimize. They don't actually break, but all seem to have different buffer-depths requiring special config to prevent overruns which lead to poor performance.

Linux and Windows will probably always beat the BSDs here, the BSDs don't go after supporting every POS hardware product on the planet, they focus on supporting the ones that get used in production environments where performance and reliability are far more important than cost. If you want to use some half assed/cheapo/bargin bin hardware product, *BSD isn't what you want, thats not its focus and anyone who uses it will tell you so.

With ZFS now being considered production ready, RAID cards are quickly becoming a whole lot less important. Presenting the raw devices to the OS and specifically the ZFS layer is a far better way to go than having the controller present a logical disk. ZFS NEEDS to know about the physical hardware directly to do its job. RAID cards which present logical drives composed of multiple physical drives to the OS are actually a bad thing.

Re:FreeBSD and i/o monitoring

[petrus4]

If you want to use some half assed/cheapo/bargin bin hardware product, *BSD isn't what you want, thats not its focus and anyone who uses it will tell you so.

This is partly true, though not as much as you might think. My hardware is probably the most disgusting, low end crap you've ever seen (no, not quite including a dead cat, but close! ;)) and FreeBSD runs on it just fine.

I'd suspect that the rule of thumb is more the age of what you've got. *New* cheap garbage won't work, but cheap garbage from probably more than two years out is pretty much guaranteed. I'm still waiting for NetBSD to release the analytical engine port. ;)

Re:FreeBSD and i/o monitoring

[micheas]

the BSDs don't go after supporting every POS hardware product on the planet, they focus on supporting the ones that get used in production environments where performance and reliability are far more important than cost. If you want to use some half assed/cheapo/bargin bin hardware product, *BSD isn't what you want, thats not its focus and anyone who uses it will tell you so.

FreeBSD supports at least commonly used low end network cards:http://fxr.watson.org/fxr/source/pci/if_rl.c

Re:For those that want to skim TFA for the bad res

[TheRaven64]

It's worth noting that one of the things in RC1 is that ZFS is now considered 'production ready' by the FreeBSD ZFS team. Maybe they should enable it for FreeBSD and then add a benchmark of shapshot and transactional I/O performance...

File system choice for databases

[shani]

ext3 is also a journaling file system. Perhaps you meant ext2?

I'm not sure you want to run fsck on an unclean shutdown on your nice big database partition either. Maybe using a journaled file system isn't such a bad idea. Also, it can be much faster for certain workloads:

http://en.oreilly.com/oscon2009/public/schedule/detail/8432
http://wiki.postgresql.org/wiki/HP_ProLiant_DL380_G5_Tuning_Guide

Note that while sequential writes could be much slower with journaled file systems, random writes were typically much faster. This is what one would expect, given how journaled file systems work.

Re:For those that want to skim TFA for the bad res

[Bralkein]

Yeah there are loads of configuration options that are relevant on servers. Going with the defaults might be sensible when comparing desktop OSs (since Mr Noobey doesn't know how to select a more suitable IO scheduler) but we've already established that they're looking at server OSs here. I think Phoronix show here that they really haven't thought this comparison through. Once the competence of a benchmarker is called into question all of a sudden I feel a massive urge to completely ignore everything they say.

It's good though that Phoronix have found a way to unite us Linux and BSD users in a joint criticism of their crappy benchmarks :)

Criticism of the test

[DaMattster]

While, interesting the test wasn't really comparing apples to apples. For example, FreeBSD RC 8 is using a slightly older version of GCC so the first test is somehwat invalidated. It is also notwworthy that several of the tests are canted towards desktop use. FreeBSD, you could argue, is not as optimized towards desktop use as Ubuntu. A better test would be to compare the Ubuntu Server edition with FreeBSD. I would like to see a test designed around the number of simultaneous website hits either OS can take or a test to see the number of concurrent ftp connections. I believe if the comparison were closer to apples and apples, we might see a more objective test.

Re:Criticism of the test

[BitZtream]

I'd agree, but I think that if you did that, FBSD would arrive as the clear winner when you're talking network IO. FBSD is known for having the absolute fastest IP stack available in a full featured OS. That costs the system in other ways, which means it doesn't cater to the desktop interactivity that an end user would want. This of course is why Ubuntu should beat the crap out of it for desktop related tasks, Ubuntu is most certainly intended to be more desktop usable than FBSD, regardless of how many people try to shoehorn FBSD onto the desktop :) I was one of those people until I could no longer take the insanity of running FBSD 4.0-CURRENT on my desktop. Of course, it was a bleeding edge code base in total upheaval as they were rewriting major chunks to be more multiprocessor friendly, it just made me realize its not a desktop OS, and of course that running -CURRENT is for people who like pain. -CURRENT hasn't been that bad since, thank god.

Re:Criticism of the test

[Zephiris]

More recent versions of FreeBSD (starting with about 5.0, of course) dramatically improved desktop interactivity with the option of using SCHED_ULE (now the default).

If you know old (in that case, ancient; pre-2000) FreeBSD, but know new Ubuntu, of course you're going to be biased in favor of whatever you use now. But comparing new Linux vs. 10 year old memories of FreeBSD isn't exactly fair.

Of course, that's also comparing an old CURRENT version, which they've never recommended any non-developer to use as their primary desktop or for production machines, and often had debugging enabled which a non-savvy person wouldn't have turned off to improve interactivity or speed.

Most of the Phoronix tests used here don't enlighten the reader at all about the characteristics of the underlying OS, but are metrics mostly based on compiler version or filesystem options.
You can tell someone's lying if Linux ever wins on 'compile speed' for most C projects, though.

Even on the aging UFS filesystem, FreeBSD has traditionally blazed there by a margin of 2-5x on both UP and SMP. On C++, it's more CPU bound, but still edged out. Even if ext4 really did speed things up (most benchmarks and independent evaluators have concluded it doesn't impact the worst cases much), you still wouldn't see it beat FreeBSD, not by such a large margin, and not 'all of a sudden', either. The problem is, the CPU scheduler is also a major indicator of how quickly anything is allowed to compile. (FreeBSD's ULE > Linux's BFS > Linux's default CFS)

Re:For those that want to skim TFA for the bad res

[Kjella]

Once the competence of a benchmarker is called into question all of a sudden I feel a massive urge to completely ignore everything they say.

You mean like pretty much every benchmark where there's a sore loser? Well, I'm sure they'll be glad to hearing spreading FUD is such an effective strategy. I'm not talking about this specific case, but for benchmarking in general your position is rather sad.

Indeed ALL instances of FreeBSD winning are assuch

[Nicolas MONNET]

All one (1) of them.

Once again, benchmarks fail

[bonch]

How about we see this against a version of FreeBSD that doesn't have debug on according to /usr/src/UPDATING?

NOTE TO PEOPLE WHO THINK THAT FreeBSD 8.x IS SLOW:
FreeBSD 8.x has many debugging features turned on, in both the kernel and userland. These features attempt to detect incorrect use of system primitives, and encourage loud failure through extra sanity checking and fail stop semantics. They also substantially impact system performance. If you want to do performance measurement, benchmarking, and optimization, you'll want to turn them off. This includes various WITNESS- related kernel options, INVARIANTS, malloc debugging flags in userland, and various verbose features in the kernel. Many developers choose to disable these features on build machines to maximize performance. (To disable malloc debugging, run ln -s aj /etc/malloc.conf.)

Re:Once again, benchmarks fail

[ground.zero.612]

Now, it may be that your karma is just shit, so if that is so please excuse this post. If not, whoever modded you down is a jackass. Someone please mod parent up!

Re:Once again, benchmarks fail

[ens0niq]

Debug? No, Sir.
1, 2

Re:Once again, benchmarks fail

[ens0niq]

Sorry, sir

From Phoronix:

"The tests that were carried out under FreeBSD 7.2, FreeBSD 8.0 RC1, and Ubuntu 9.10 Alpha 6 included timed ImageMagick compilation"

From freebsd-current mailing list:

> Hi,
>
> I would like to ask that the FreeBSD 8.0-RC1 ISO-s free from the
> debugging features (WITNESS, malloc debugging, etc.)? Or these
> services are still being active?

They are gone, for the most part. r197065: http://svn.freebsd.org/viewvc/base?view=revision&revision=197065

Remove extra debugging support that is turned on for head but turned off for stable branches:

- shift to MALLOC_PRODUCTION
- turn off automatic crash dumps
- Remove kernel debuggers, INVARIANTS*[1], WITNESS* from
GENERIC kernel config files[2]


[1] INVARIANTS* left on for ia64 by request marcel
[2] sun4v was left as-is

Re:For those that want to skim TFA for the bad res

[Bralkein]

Perhaps "called into question" wasn't the best choice of phrase. I certainly don't mean that if absolutely anyone, anywhere cries BS about a benchmark then the benchmark is instantly void, rather I mean that once a benchmarker starts to look like they don't know what they're doing then since I don't have enough time to pick through every inch of their technique and raw data, it's easier just to file the whole thing under "useless" in my brain and move on with my life.

The problem is that there are so many poorly-conducted benchmarks around that really end up saying nothing that it can make one jaded and perhaps quicker to reject a benchmark than would be proper. Good benchmarks are certainly worth their weight in gold (erm, or maybe more. I don't know how much a benchmark weighs!).

Im colourblind you insensitive clod

[masmullin]

NEVER solely use colours to describe things.

Re:Im colourblind you insensitive clod

[Korin43]

Perhaps they should use colors in combination with something else, like say.. bars and lines? We could show the difference based on distance from the bottom line.

Re:Im colourblind you insensitive clod

[Just Some Guy]

The current ones are grey. Trust me: you wouldn't be missing anything.

Re:GPLv3 stuff is not allowed in the base system

[notamisfit]

Certain companies (Apple) have no-GPLv3 policies in place.

Re:Sigh...

[Just Some Guy]

freebsd is 100% binary compatible with linux.

Snicker. I love FreeBSD and run it on all the servers I administer, but the Linux compatibility stuff pretty much ends at usermode. Good luck firing up VMWare or anything else that requires an un-ported kernel module! In practice, every program I personally want to run on FreeBSD is available as a native binary or in ports, with the exception of programs that require kernel mods, in which case they won't work at all anyway.

Debugging/self-check in the FreeBSD kernel

[cracauer]

People mentioned the self-checks and debugging features that used to be turned on in FreeBSD development branches and beta releases.

Self-checks, which are the major source of kernel slowdowns in those kernel options, are not turned on in the 8.0 release candidates.

Debugging is on, but unless you are very short of memory it should not cause a noticeable slowdown.

FreeBSD's slowness in these benchmarks can be attributed to two factors:

1) the compiler. The GPL v3 is unacceptable for FreeBSD, so newest GCCs cannot be used as the base compiler. Users can choose to install a new gcc on their own (as a port) and then even go and compile all ports and even the base system with that new compiler (some parts might not have been cleaned up to comply with new language strictness that might have come with new gccs).

2) threading, as in the userland threading library support. It is very hard to tell whether there is some performance problem in FreeBSD's thread libraries or whether applications just happen to have been optimized and tested only with Linux.

That happens a lot and you can also see Solaris with it's M:N threading fail miserably for some threading benchmarks that do dump things, such as just creating and destroying threads at a high rate.

%%

The problem of "thread benchmarks" benchmarking bogus things and/or just having been written with Linux' thread model in mind has been going on for 12 something years now. Benchmarking thread systems in a realistic manner is very difficult. In real-world applications you don't create and destroy threads at a high rate and you minimize locking. Benchmarking this is almost as hard as benchmarking programming languages.

I haven't benchmarked threading libraries, knowing that it will take time that I don't have right now. I can tell you, however, that just the I/O subsystem in FreeBSD, as in filesystems and networking, isn't any slower than Linux. Not to mention GbE and today's disks are too slow to really show an OS difference for most tests anyway.

%%

The real question of I/O subsystems will come in when ZFS+Zraid is standard in FreeBSD and BTRFS is standard in Linux. In a couple of years from now nobody will understand why we ran today with no snapshots, with the raid hole (from block layer raid systems) and without transparent compression in the subtrees where we want it.

But these filesystems are complicated and there's some real performance difference visible.

%%

An area completely overlooked in the benchmark is the VM subsystem. Namely - what happens when you overload your RAM and paging commences? Linux used to make very bad choices here (dropping readonly pages, which is a wise thing as such, at rates about 10 times higher than wise).

FreeBSD used to be the go-to OS on memory shortage thanks to John Dyson's VM work (backed by a large database company that provided support and a realistic benchmarking environment during that development).

But today? Nobody knows. I'm not aware of any benchmarks that you can download that simulate memory stress and map the tradeoffs that the OS makes.

In general, the biggest obstacle to improve Linux, FreeBSD and everybody's else's OS performance is a lack of high-quality benchmarks.

Why don't people develop more benchmarks? Because they get annoyed that today, in 2009, no realistic OS benchmark will show a single number as a result. All OS benchmarks today can only make a map of what tradeoffs the OS chose, what part of the running application suite got the short end in favor of what other part. This isn't sexy and publishers don't like it.

People like reality reduced to single numbers, but in the area os OS benchmarks (and language benchmarks) that party is over.

Myself, I found myself gasping many years ago when I benchmarked network I/O versus userland CPU load. I hammered a couple of GbE interfaces while at the same time running moderate memory-intensive CPU benchmarks (with no network access from those CPU lo

Re:Debugging/self-check in the FreeBSD kernel

[petrus4]

But today? Nobody knows. I'm not aware of any benchmarks that you can download that simulate memory stress and map the tradeoffs that the OS makes.

In general, the biggest obstacle to improve Linux, FreeBSD and everybody's else's OS performance is a lack of high-quality benchmarks.

The call for technical integrity here, and the use of %% as a seperator, implies to me that the author of the OP is (at least potentially) a member of the old school.

It causes me joy to know that there are still some of you left, and even more, that there are still some left here on Slashdot.

To both the FSF and Mark Shuttleworth, I can only say, do your worst. UNIX will survive.

Re:Uptime?

[petrus4]

Let's try uptime? BSD vs Debian. No contest. When you can measure a debian servers uptime in years, not weeks, let me know.

Only if ctrl-alt-backspace zapping of X is re-enabled. I have a bad 3d card, (not FreeBSD's fault) with corrupt video ram, and it was causing X to lock up. Before I re-enabled the key combo, I had several instances where I had to reboot. *sigh*

Re:For those that want to skim TFA for the bad res

[saleenS281]

Which is why one has to wonder why they used bleeding edge ext4 on Linux, and didn't use ZFS on FreeBSD. It couldn't be because they were trying to intentionally skew the numbers! And yes, ZFS is fully baked on FreeBSD.

Re:Sigh...

[jedidiah]

>> You can only do an equally good job with MacOS.
>>
>> For this grant priveledge you get a meagre selection of hardware that will
>> either gouge you or leave you wanting.
>>
> Feeding the trolls with more trolling? My 2009 Macbook "white" is no more expensive than PC
> laptops in the same form factor. Actually it's a bit cheaper than most comparable 13" laptops
> w/ a faster FSB than most.

In the same form factor? If you are going to argue this stuff at least make a little
sense when you do it. Mac laptops compete favorably with PC laptops when you include
"similar components".

Although even that proves my point: "Purevideo or nothing" is the Mac option.

>
> I have no performance complaints and for an IGP the GeForce 9400M certainly doesn't suck.
> ...and it's your ONLY OPTION

Widen the field to laptops with anything else and the Mac gets creamed.

>
> And, I don't have to deal with the crappy PC BIOS anymore.
>

Like that is a real problem for a user like you...

OTOH, my ION nettop with a "normal BIOS" is booted up before a mini gets to it's white screen.

If Macs represent the state of EFI, then give me a small boot disk so I can avoid it.

Debug?

[nurb432]

Doesn't the kernel code in 8RC1 still have debugging stuff enabled? I know that is a speed killer and would at the least skew the numbers.

Re:Sigh...

[Xtifr]

Learning to use focus-follows-mouse isn't hard, though I admit it can be a little confusing at first; going back to crappy systems that don't offer it after you've learned its benefits is what's hard!

Re:Sigh...

[Wolfrider]

--Nice try, but IMHO it's not " 100% " until you can run Vmware Workstation on FreeBSD.

The Linux scheduler historically isn't thread-fair

[tlambert]

The Linux scheduler historically isn't thread-fair

Specifically, as long as there is work to do in the thread, it doesn't force a context switch; this attribute is why the GIF renderer on Netscape would crash on FreeBSD but not on Linux, because after an involuntary preemption, FreeBSD did not guarantee that the preempted thread would be the one reselected to run; Linux did, but in the process eliminated scheduling fairness for the other threads. So FreeBSD defaults to better thread fairness. This is also why the SQL benchmarks from last year were better on Mac OS X 10.5.5 (Leopard) than on Ubuntu.

If Linux actually honors the SCHED_RR policy, and were to explicitly set it, it would probably do better on the SQL benchmarks; however, it would likely do worse than with SCHED_OTHER or SCHED_FIFO (not sure which is the implementation default in the current Linux scheduler), but certain workloads like to serialize their operations and get better cache line efficiencies from doing so. Oh, and if the default scoping is PTHREAD_SCOPE_SYSTEM vs. PTHREAD_SCOPE_PROCESS will also matter here (again, assuming it's supported).

A decent benchmark would explicitly set the policy and scope to the best value for the system it was running on. The FreeBSD numbers could have easily been better as well, by adjusting these values to explicit values, rather than taking the system defaults.

-- Terry

Look at the little triangle ...

[shis-ka-bob]

The triangle in the upper left corner points up when up is good (operations per second). It points down when down is good ( seconds to complete test suite). Seems pretty clear to me.

Re:Sigh...

[Timothy Brownawell]

Focus follows mouse is available as an option in Windows. It has been since the first release of PowerToys for Win2k.

I've tried that (well, the XP version), and last I checked there was no way to turn off raise-on-click. Which makes it mostly useless.

Re:For those that want to skim TFA for the bad res

[badkarmadayaccount]

Unless the kernel does some batshit insane OTF code generation optimizations.