FreeBSD/Alpha SMP fully multiuser stable and checked in 27
David O'Brien writes "
FreeBSD SMP on Alpha processors is now stable and has been checked into
CVS repository on the development branch (5.0-CURRENT). John Baldwin, Andrew Gallatin, and Doug Rabson first booted single-user SMP on Friday, April 13th, and full multi-user support followed on Tuesday, April 17th. The code was stablized and then checked into the CVS repository on Friday, April 27th. Since being checking into the repository, it has been verified to work on quad-CPU DEC AlphaServer 4100s, dual-CPU Compaq DS-20s, dual-CPU API UP2000, and dual-CPU DEC AlphaServer 2100s.
"
Re:great news for FreeBSD server farms (Score:1)
Lets see. 4.3 was scheduled for April 15,2001. It was released formally on the 20th of April.
5 days off. Much better than target dates on Pink or Rhaposdy (apple), Windows (microsoft), or how about the year "late" the linux kernel was.
4.4 FreeBSD is due out in August. Care to put your money where your fat mouth is Josh? How does $100 sound? I take $100 of your money if 4.4 is not released in August. (I'll then send your money to Jordan so he can buy a round of doughnuts for the office) Just sell some of your underwater Linux IPO stocks.
Re:Now, to get AMD to admit BSD exists. (Score:1)
Another K6 owner has the same problem. I took up the K6 2.9V 200 and put in a Pentium 166 and the problem went away. He stopped buying AMD.
Re:Good Thing (Score:1)
Re:great news for FreeBSD server farms (Score:1)
Excuse me, but I believe you forgot to enclose your comment in <moron> tags. Please don't let it happen in the future.
the evidence [byte.com]
Re:great news for FreeBSD server farms (Score:1)
> It's not Linux thats wiping the floor, its TUX; the tiny in-kernel httpd that is designed for TRIVIAL matters. Putting the daemon inkernel is a glaring risk. It doesn't match the completeness of other userland servers like Apache.
First, TUX pushed the kernel to be fast in general. Somebody recently announced a user space web server called X15 on the linux-kernel mailing list that is more-or-less the same speed as TUX.Second, with respect to the comment about not "matching the completeness" of Apache. Both TUX and X15 support CGIs, and I know that TUX has a plug-in model (I wouldn't be surprised if there is a plug-in model in X15 either.) If there is a Apache module that you like, but you need a faster web server than Apache, maybe you should port the plug-in...
Finally, SpecWeb is a very real test of OS and web server performance. Something like 30% of it's output is driven by CGIs. It's not a trivial setup. Not only does it stress the networking stack, but it stresses the VM system and disk/block/vfs layer very hard too. (The working set of data on the web server can be larger than the amount of memory installed in the system.)
I know this sounds like flame bait, but I'm sick of the FreeBSD advocates claiming that they are faster than Linux when it comes to network servers. If that's true, where are the numbers to prove it?
I want to see SpecWeb results!
Re:Not soon enough... (Score:1)
X11 on OBSD (Score:1)
How could XFree86 4.0.3 be more updated?
Re:great news for FreeBSD server farms (Score:1)
a VERY STUPID IDEA...
ok, so you gain some speed, at what cost? instability? security? why do you think IIS has such fucked exploits?...
badly written scripts become MUCH more dangerous in such a situation...
Thru, threading support needs some work, but there are many test that show that on OS's like the BSD's and linux, forking is actually a more efficient method than threading.. if i had a heavy threading app on mulitple cpu's i would look no further than solaris...
Choose the OS base on merits, not on zealoutry...
FreeBSD + Athlon SMP == working... (Score:1)
Dual Athlon 1.2GHZ CPU's, FreeBSD SMP, stock -current branch.
Re:Now, to get AMD to admit BSD exists. (Score:1)
oh, yeah, cuz NT sucks donkey dicks on commodity hardware.
Re:great news for FreeBSD server farms (Score:1)
a VERY STUPID IDEA..."
You obviously haven't even researched TUX to begin with. Typical BSD zealot.
CGI applications can be configured as user-level modules in TUX. TUX is _VERY_ safe AND stable. In fact just as safe as in-kernel firewalling. What, BSD doesn't have in-kernel firewalling or NFS because they introduce stability or security problems? Or that too, a "VERY STUPID IDEA"? Even OpenBSD has kernel space server code.
"why do you think IIS
has such fucked exploits?."
Because IIS was designed for a different OS (with all it's problems), is a very different server than TUX and has been written by a company that doesn't care about security among other reasons.
"many test that show that on OS's like the BSD's and linux, forking is actually a more efficient method than threading.."
Give me at least a single URL please.
"...if i had a heavy threading app on mulitple cpu's i would look
no further than solaris..."
There are many kinds of thread libraries out there (many not kernel-dependent), and you do not have to have an SMP machine to enjoy benefits of threading. Duplicating resources per every process is very expensive, not mentioning the terrible cost of process start and context switch compared to just a thread.
In fact take a look at http://oss.sgi.com/projects/state-threads.
Anyway going back to my original post, let's compare BSD and Linux running a SPED server like Zeus or something like a TUX server and see the results. Get your facts first before flaming.
Re:great news for FreeBSD server farms (Score:1)
Re:Now, to get AMD to admit BSD exists. (Score:1)
After the 233 continued to sig 11 during kernel builds (Freebsd 4.2-release), I put it on a new soyo 5ehm, and it did the same thing. I replaced the 233 with a k6-2 350 on the soyo, and it works fine.
Re:Now, to get AMD to admit BSD exists. (Score:1)
---------------
Re:Good Thing (Score:1)
All I want to know is when OpenBSD is going to get SMP for any platform...
Usually, after NetBSD gets it running just fine, they will port it.
--
Good Thing (Score:1)
All I want to know is when OpenBSD is going to get SMP for any platform...
Re:great news for FreeBSD server farms (Score:2)
I don't have specweb results. I don't care about specweb. Further, I do not agree with you that it is an adequate test of network serving performance.
Here is a real-life, well set-up and functioning test - posted on slashdot weeks ago. Please note that the "primitave" FreeBSD 4.2 SMP _still_ outperforms linux 2.4 - and it's going to be disgusting about 3 months after 5.0-RELEASE, when the thing matures.
benchmark here [byte.com]
If you don't believe the results, set up a test like that on your own hardware, and experiance the fact that linux is playing "follow the leader" first hand.
Re:great news for FreeBSD server farms (Score:2)
You prefer a byte.com self described opinion column over industry standard benchmarks...
No comment.
Re:great news for FreeBSD server farms (Score:2)
We run probably the second biggest cluster of FreeBSD servers in the world, second to yahoo. And trust me, in our internal test, with standard 2cpu, 2HD servers, the load that a Customized FreeBSD install could handle was anywhere from 1.3-2 times the load a 2.4 machine could under stress.
And this was using plain old FFS and no softupdates. Even with linux's stupid async mounts, FreeBSD still mopped the floor.
no specweb benchmarks on FreeBSD mean nothing. To those who know, it will still be the performance champion for load intensive and clusterable applications.
I talk from experience, not out of my ass like you.
My opinions are my opinions, deal with it!
Great News (Score:2)
1. FreeBSD has had excellent SMP support since 4.0.
2. The SMP subsystem has been rewritten to make it even more efficient.
3. Arguing that FreeBSD doesn't run well on the Alpha is like arguing that Linux does not run well on the DreamCast. The Alpha release is still in development and there are still some problems with it. (Check out DEC hardware sometime... It is definitely not the nicest platform to develop for).
The simple fact is that some of us like FreeBSD a lot more than Linux. The fact that we have actually run both operating systems for extensive periods of time, and you have not, really should not matter should it?
The day linux has anything as sensible as the ports tree, softupdates, cvsup, make world, etc. give me a call. Until then I am not interested.
I, for one, got tired of moving from one linux system to another and having to search all over to figure out where a particular distribution put all of the critical files. For example, does it use sys-v style rc.d runlevel scripts or does it use BSD style rc scripts? Where does it stick particular programs? Slackware is as different from Debian as BSD is from Solaris.
Also, ever note how the FreeBSD folks seem to have all the cool drivers before the Linux people? Why do you think that is? FreeBSD had working I2O support in 3.5. Linux _still_ does not have a sensible I2O subsystem. FreeBSD had USB way before Linux did.
Not to mention, all of the projects the FreeBSD team is working on always seem to grow out of research papers, whereas the Linux projects seem to lack the same direction
-sirket
Industry standard benchmarks (Score:2)
And, there was much whining and knashing of teeth over their results.
Still no comment?
What the hell are you talkin' about! (Score:2)
Funny story... Windows wont even run on my k6-2 400 without being patched. It doesnt correctly share interrupts on the FIC-2013+ motherboard. Guess what... Linux, FreeBSD and OpenBSD cause no problem. I was so proud that I found a motherboard that wasnt 'supported' by Win98.
Not soon enough... (Score:2)
Re:Now, to get AMD to admit BSD exists. (Score:3)
My guess is that your motherboard is either bad quality, or perhaps needing a decent BIOS update. Another candidate could be your RAM. I once had problems with an old Asus SC-200 SCSI controller, that was not up to the task with faster systems anymore. Overclocking is a bad idea as well.
Re:great news for FreeBSD server farms (Score:3)
Re:Now, to get AMD to admit BSD exists. (Score:4)
Considering I develop and maintain the GCC/Binutils toolchain on AMD K6 machines, I can say absolutely this is false. You may easily have a HW problem. I would bring this up in one of the FreeBSD mailing lists to help determine what your problem is. I will note, I had to increase the voltage on one of my K6-2/450 machines by 0.1v, other wise it would SIG 11 as you mention. `make world' really stresses a machine and if your mobo just isn't keeping the volage up, you will have problems.
-- David
Re:great news for FreeBSD server farms (Score:5)
No code is being replaced. They're not merging the BSD/OS 5.0 experimental kernel at all; SMPng is based off of some of the design concepts in it.
If FreeBDS 5.0 really does get an "entirely overhauled SMP structure" in six months, then it will be just that: immature code that's only been around a few months.
*BSD tests it first before releasing it. It makes more sense then to throw it to the public. When was the last time a *BSD had an ontime release?
If FreeBSD's TCP/IP is so kewl, then why doesn't anyone submit SPECweb benchmarks using FreeBSD for the 1 CPU category? Linux wipes the floor with Solaris, AIX, Tru64, and Windows 2000.
It's not Linux thats wiping the floor, its TUX; the tiny in-kernel httpd that is designed for TRIVIAL matters. Putting the daemon inkernel is a glaring risk. It doesn't match the completeness of other userland servers like Apache.