SGI publishes their code. It's just that their changes are so radical and "dirty" that they're not useful/mainteinable for the rest. Remember, SGI has sold 256 CPU machines with their 2.4 kernel - where 2.4 vanilla doesn't works very well beyond 8 cpus
I don't think many people realize how _HUGE_ is a 32 processor box (myself at least I can't even imagine it). As someone already told you, win 2k3 supports 64 CPUs not 32. But I bet they could make it run in ej: 128 cpus with no much effort (if any).
Why they don't to it? Plain simple: money. That SGI box is a special case because some "national" thing. Enterprises don't use 1024, or even 128 or 64 cpus in most of the cases: The run clusters of small machines, that's what it sells today and that's where Microsoft is putting their eyes.
That sounds like the "Mount paradigm": To increase speed you can do wirte operations asynchronous. So you have to "umount" things with the lettle icon.
I guess it'd be much easier to "mount" things as sync. That way you wouldn't have to use that icon, would you?
There hasn't been a lot of releases lately; I've been searching and I was wondering if Minimo would be a suitable replacement for Mozilla in 486-pentium boxes...
Actually there's a better tool: frag (in defrag debian package)
frag -s/etc summary:
2% file fragmentation (66 of 2903 files contain fragments)
20% block fragmentation (3056 of 15241 blocks are in fragments)
0% overall fragmentation (73 fragments out of 15241 blocks)
Average inter-fragment gap length = 4616
frag -s/var summary:
4% file fragmentation (560 of 11432 files contain fragments)
46% block fragmentation (149948 of 324734 blocks are in fragments)
0% overall fragmentation (945 fragments out of 324734 blocks)
Average inter-fragment gap length = 55889
And BTW; Windows XP _defragments_ the filesystem, so that means you can correct fragmentation. Unix filesystems have good algorithms for this but what happens if the filesystem changes? If you deleted a file the decisions made by those algorithms are no longer valid. (Windows XP actually recolects information of the programs more frecuently used and defragments automatically the filesystem according those data. So the apps you use more often are arranged in a way they load faster, ie: you put the shared libraries near the data files some app is going to use, so the disk loads it faster instead of seeking over the disk. Theres nothing like that under *nix i think)
He benchmarked 2.4, which these days is a bit...pointless. Ext3 has a lot of improvements only in 2.6, just like the other FSs. 2.4 filesystems are in "mainteinance mode" where in 2.6 there's been a large amount of development. If he'd run 2.6 he could benchmark reiser4 too.
No. (and I'm from spain too). Telefonica is much worse. Telefonica (for those who don't know) was created by our "beloved" Franco dictator. It's been a "public company" for years and there was no other company on the country. In 1996 The popular party won the elections and they "privatized" several public companies (which was the right thing to do). However, they did it *so* wrong that Telefonica owns everything, and the other few telecommunications companies just rent Telefonica's infrastructures because it's not worth wasting money in creating their own infrastructure.
Which means that Telefonica is now a private company which controls everything, our ADSL lines are one of the slowest and of the more expensive on the whole EU, if not the worse. Latencies are huge because Telefonica has disabled the "FAST PATH" option in their rowters, they can't do anything since their infrastructure couldn't handle good latencies, and there's not a lot of incentive to fix that with new infrastructures because there's no competence.
"As Gentoo users are generally an annoying nusiance in my IRC channel, I would like to be able to set a +b on *!gentoouser@*, in order to keep them out. In order to accomplish this, I would like Gentoo to set the default username to something Gentoo specific.
Reproducible: Always
Steps to Reproduce:
1. Join an IRC channel
2. Wait for a Gentoo user to come in, talk about how he "emerged" this and that and all the "optimizations" he used, but something still isn't working
3. Mention that he should have read the FAQ before/joining and/kick him."
I wish I could do the same in my channels. I'm tired of hearding "gentoo is 200% faster" "It eats less memory" (how on earth can optimizations make a program eat a noticeable amount of memory - if anything they're usually bigger because of the loop unrolling). They never show numbers though.
Well, being me a dialup linux user...
Yes, I like dialup. And yes, I've tried 2Mbit connections for several months. First of all, broadband in spain is painful slow and expensive. And no, I'm not exagerating. broadband average latency isn't much better than the one froma dialup for example, because Telefonica/terra odesn't want to switch on some "FASTH PATH" option in their router (existing lines can't handle it)
First, it's cheaper. Period
Second, I got constant 5KB/s. I use a up-to-date debian sid, and I download some isos if i left it switched on in the night. Bandwith is more than enought for web surfing.
Third, I move weekly to different houses, dialup allows me to be charged *one* time when I move, they charge me the same regardless where I call from.
Fourth, I've no access to any broadband service *at all* in my small village.
So no, I'm not going with broadband. Sorry.
ketchup is a script that automatically patches between kernel
versions, downloading and caching patches as needed, and automatically
determining the latest versions of several trees. Example usage:
$ ketchup 2.6-mm
2.6.3-rc1-mm1 -> 2.6.5-mm4
Applying 2.6.3-rc1-mm1.bz2 -R
Applying patch-2.6.3-rc1.bz2 -R
Applying patch-2.6.3.bz2
Applying patch-2.6.4.bz2
Applying patch-2.6.5.bz2
Downloading 2.6.5-mm4.bz2
Downloading 2.6.5-mm4.bz2.sign
Verifying signature...
gpg: Signature made Sat Apr 10 21:55:36 2004 CDT using DSA key ID 517D0F0E
gpg: Good signature from "Linux Kernel Archives Verification Key
"
gpg: aka "Linux Kernel Archives Verification Key
"
owner.
gpg: WARNING: This key is not certified with a trusted signature!
gpg: There is no indication that the signature belongs to the
Primary key fingerprint: C75D C40A 11D7 AF88 9981 ED5B C86B A06A 517D 0F0E
Applying 2.6.5-mm4.bz2
uh? that's just not true. In linux if you didn't enabled acpi and you want to enable it you'll have to recompile the kernel. But then you *shouldn't* have disabled it (the fact that you can disable it is that that opcion is for experts, you DO NOT have to touch that). And you can load as modules almost any driver. Join that to a hardware detector and you'll find that well, linux works as windows, loading/unloading modules where neccesary
Sure it has antialiased font. I just find weird that that guy switched them off when doing screenshots. I could get antialiased fonts in windows 95 + IE 4.0
GREAT article (I never though KDE could have something so cool like thumbnails of the files when you're moving files and you're going to overwrite some)
Now, If he could have used a nice theme (keramic is enought) and antialiased fonts it'd perfect....
Yes, It's done. Ssearch for "CFQ scheduler". It's in the -mm tree. You've "io priorities" so you can tell apache "you've been guaranteed 80% of the disk bandwith". Or run updatedb cron jobs at a lower io priority so they don't interfere with mozilla/openoffice...
Irix has it already for years. Linux has it now. MS is planning this for Longhorn...(it makes the OS a bit more "realtime" so you won't have video pauses because your videos have a highter priority...)
1000x is not a inveted number, just run "find/" while you're running "cp/dev/zero/tmp" (all on the same disk). Measure how many time takes in 2.4 and how many time it takes in 2.6 with anticipatory scheduler.
Nobody has mentioned, BTW, the CFQ scheduler. CFQ is the coolest thing you'll find for desktops. AIUI, you've a list of queues for each process, and you just round robin on them. This provides *extreme* fairness for desktops.
In fact, Jens Axboe used that to provide "io priorities". That means you can give "ionice -20" to a process and that process will have more disk bandwith than other processes. Or, in other words, you can run your updatedb cron jobs with a lower io priority. So you won't notice it while you're using openoffice/mozilla because the updatedb cron job is getting less io priority...
This makes linux a bit more "realtime OS", since we'll be able to say "I want that apache has been guaranteed 80% of the disk bandwith". And that will be true _always_, no matter if your users are doing stupid things....
Slashdot Mirror
SGI publishes their code. It's just that their changes are so radical and "dirty" that they're not useful/mainteinable for the rest. Remember, SGI has sold 256 CPU machines with their 2.4 kernel - where 2.4 vanilla doesn't works very well beyond 8 cpus
Linux _DOES_ run in those machines with vanilla kernels. I guess throw their own patches but there's no doubt vanilla kernel is good enought: http://marc.theaimsgroup.com/?l=linux-kernel&m=108 341362028320&w=2
I don't think many people realize how _HUGE_ is a 32 processor box (myself at least I can't even imagine it). As someone already told you, win 2k3 supports 64 CPUs not 32. But I bet they could make it run in ej: 128 cpus with no much effort (if any).
Why they don't to it? Plain simple: money. That SGI box is a special case because some "national" thing. Enterprises don't use 1024, or even 128 or 64 cpus in most of the cases: The run clusters of small machines, that's what it sells today and that's where Microsoft is putting their eyes.
Fact: You're idiot. Any questions?
That sounds like the "Mount paradigm": To increase speed you can do wirte operations asynchronous. So you have to "umount" things with the lettle icon. I guess it'd be much easier to "mount" things as sync. That way you wouldn't have to use that icon, would you?
There hasn't been a lot of releases lately; I've been searching and I was wondering if Minimo would be a suitable replacement for Mozilla in 486-pentium boxes...
Actually there's a better tool: frag (in defrag debian package)
/etc
/var
frag -s
summary:
2% file fragmentation (66 of 2903 files contain fragments)
20% block fragmentation (3056 of 15241 blocks are in fragments)
0% overall fragmentation (73 fragments out of 15241 blocks)
Average inter-fragment gap length = 4616
frag -s
summary:
4% file fragmentation (560 of 11432 files contain fragments)
46% block fragmentation (149948 of 324734 blocks are in fragments)
0% overall fragmentation (945 fragments out of 324734 blocks)
Average inter-fragment gap length = 55889
And BTW; Windows XP _defragments_ the filesystem, so that means you can correct fragmentation. Unix filesystems have good algorithms for this but what happens if the filesystem changes? If you deleted a file the decisions made by those algorithms are no longer valid. (Windows XP actually recolects information of the programs more frecuently used and defragments automatically the filesystem according those data. So the apps you use more often are arranged in a way they load faster, ie: you put the shared libraries near the data files some app is going to use, so the disk loads it faster instead of seeking over the disk. Theres nothing like that under *nix i think)
He benchmarked 2.4, which these days is a bit...pointless. Ext3 has a lot of improvements only in 2.6, just like the other FSs. 2.4 filesystems are in "mainteinance mode" where in 2.6 there's been a large amount of development. If he'd run 2.6 he could benchmark reiser4 too.
are you offering? 8) AFAIK this is *NOT* a final release and they haven't said "WE'RE NOT SUPPORTING BSD"
Linux has always had lots of VM tunables, ie: /proc/sys/vm/bdflush etc. swappiness is just a abstraction.
The government of Spain does *NOT* run TDE. IT was a public company but it was privatized. No true anymore.
No. (and I'm from spain too). Telefonica is much worse. Telefonica (for those who don't know) was created by our "beloved" Franco dictator. It's been a "public company" for years and there was no other company on the country. In 1996 The popular party won the elections and they "privatized" several public companies (which was the right thing to do). However, they did it *so* wrong that Telefonica owns everything, and the other few telecommunications companies just rent Telefonica's infrastructures because it's not worth wasting money in creating their own infrastructure.
Which means that Telefonica is now a private company which controls everything, our ADSL lines are one of the slowest and of the more expensive on the whole EU, if not the worse. Latencies are huge because Telefonica has disabled the "FAST PATH" option in their rowters, they can't do anything since their infrastructure couldn't handle good latencies, and there's not a lot of incentive to fix that with new infrastructures because there's no competence.
...I live in Spain and I use Telefonica's services. Hopefully this WILL force them to do SOMETHING after several years asking them to do that.
"As Gentoo users are generally an annoying nusiance in my IRC channel, I would like to be able to set a +b on *!gentoouser@*, in order to keep them out. In order to accomplish this, I would like Gentoo to set the default username to something Gentoo specific. Reproducible: Always Steps to Reproduce: 1. Join an IRC channel 2. Wait for a Gentoo user to come in, talk about how he "emerged" this and that and all the "optimizations" he used, but something still isn't working 3. Mention that he should have read the FAQ before /joining and /kick him."
I wish I could do the same in my channels. I'm tired of hearding "gentoo is 200% faster" "It eats less memory" (how on earth can optimizations make a program eat a noticeable amount of memory - if anything they're usually bigger because of the loop unrolling). They never show numbers though.
Well, being me a dialup linux user... Yes, I like dialup. And yes, I've tried 2Mbit connections for several months. First of all, broadband in spain is painful slow and expensive. And no, I'm not exagerating. broadband average latency isn't much better than the one froma dialup for example, because Telefonica/terra odesn't want to switch on some "FASTH PATH" option in their router (existing lines can't handle it) First, it's cheaper. Period Second, I got constant 5KB/s. I use a up-to-date debian sid, and I download some isos if i left it switched on in the night. Bandwith is more than enought for web surfing. Third, I move weekly to different houses, dialup allows me to be charged *one* time when I move, they charge me the same regardless where I call from. Fourth, I've no access to any broadband service *at all* in my small village. So no, I'm not going with broadband. Sorry.
Already saw ketchup?
http://www.selenic.com/ketchup/ketchup-0.5 :
ketchup is a script that automatically patches between kernel
versions, downloading and caching patches as needed, and automatically
determining the latest versions of several trees. Example usage:
$ ketchup 2.6-mm
2.6.3-rc1-mm1 -> 2.6.5-mm4
Applying 2.6.3-rc1-mm1.bz2 -R
Applying patch-2.6.3-rc1.bz2 -R
Applying patch-2.6.3.bz2
Applying patch-2.6.4.bz2
Applying patch-2.6.5.bz2
Downloading 2.6.5-mm4.bz2
Downloading 2.6.5-mm4.bz2.sign
Verifying signature...
gpg: Signature made Sat Apr 10 21:55:36 2004 CDT using DSA key ID 517D0F0E gpg: Good signature from "Linux Kernel Archives Verification Key "
gpg: aka "Linux Kernel Archives Verification Key "
owner.
gpg: WARNING: This key is not certified with a trusted signature!
gpg: There is no indication that the signature belongs to the Primary key fingerprint: C75D C40A 11D7 AF88 9981 ED5B C86B A06A 517D 0F0E
Applying 2.6.5-mm4.bz2
What about the time hotspot wasted to determine that that optimizacion was necesary? I'm sure it wasn't free...
uh? that's just not true. In linux if you didn't enabled acpi and you want to enable it you'll have to recompile the kernel. But then you *shouldn't* have disabled it (the fact that you can disable it is that that opcion is for experts, you DO NOT have to touch that). And you can load as modules almost any driver. Join that to a hardware detector and you'll find that well, linux works as windows, loading/unloading modules where neccesary
Sure it has antialiased font. I just find weird that that guy switched them off when doing screenshots. I could get antialiased fonts in windows 95 + IE 4.0
GREAT article (I never though KDE could have something so cool like thumbnails of the files when you're moving files and you're going to overwrite some) Now, If he could have used a nice theme (keramic is enought) and antialiased fonts it'd perfect....
The X.org CVS tree is hosted in the freedesktop.org server. So I'd expect lots of colaborations..
Yes - it's called CFQ. You can say "this process will have 35% of the disk bandwith guaranteed". It's ion the -mm tree.
Those are some benchmarks posted by the developers, developers, developers....
n ch+group:fa.linux.kernel&hl=en&lr=&ie=UTF-8&group= fa.linux.kernel&selm=fa.citnj0l.th62p7%40ifi.uio.n o&rnum=7
n ch+group:fa.linux.kernel&hl=en&lr=&ie=UTF-8&group= fa.linux.kernel&selm=fa.cjtrj8p.vh22h3%40ifi.uio.n o&rnum=1
n ch+group:fa.linux.kernel&hl=en&lr=&ie=UTF-8&group= fa.linux.kernel&selm=fa.cithjp0.thc0h6%40ifi.uio.n o&rnum=4
http://groups.google.com/groups?q=io+scheduler+be
http://groups.google.com/groups?q=io+scheduler+be
http://groups.google.com/groups?q=io+scheduler+be
Yes, It's done. Ssearch for "CFQ scheduler". It's in the -mm tree. You've "io priorities" so you can tell apache "you've been guaranteed 80% of the disk bandwith". Or run updatedb cron jobs at a lower io priority so they don't interfere with mozilla/openoffice... Irix has it already for years. Linux has it now. MS is planning this for Longhorn...(it makes the OS a bit more "realtime" so you won't have video pauses because your videos have a highter priority...)
1000x is not a inveted number, just run "find /" while you're running "cp /dev/zero /tmp" (all on the same disk). Measure how many time takes in 2.4 and how many time it takes in 2.6 with anticipatory scheduler.
Nobody has mentioned, BTW, the CFQ scheduler. CFQ is the coolest thing you'll find for desktops. AIUI, you've a list of queues for each process, and you just round robin on them. This provides *extreme* fairness for desktops.
In fact, Jens Axboe used that to provide "io priorities". That means you can give "ionice -20" to a process and that process will have more disk bandwith than other processes. Or, in other words, you can run your updatedb cron jobs with a lower io priority. So you won't notice it while you're using openoffice/mozilla because the updatedb cron job is getting less io priority...
This makes linux a bit more "realtime OS", since we'll be able to say "I want that apache has been guaranteed 80% of the disk bandwith". And that will be true _always_, no matter if your users are doing stupid things....