Dude, big iron is not comparable in the slightest to that dinky little dual PPro Linux 'server' you keep in your closet. A CPU can fail, on a live running system, and the machine and Solaris or AIX won't even hiccup.
This is NOT TRUE. Solaris will CRASH.
Your application will notice, because suddenly a couple of its threads will quit, but that's ok, software like Oracle already knows how to deal with failed transations.
If Oracle or Sybase or any other application I've EVER SEEN has one of its thread crash the WHOLE THING IS FUCKED. What if that thread held a lock? What you say is hopelessly stupid.
And if you can schedule a CPU board removal/swap, then there won't be ANY problems at all, as the OS will migrate threads to other CPUs and allow the removal or hardware.
There's a big difference between disabling the hardware and then removing it, and having the hardware fail. Once it has failed there has ALREADY BEEN MEMORY CORRUPTION and the system *MUST* PANIC AND REBOOT or it only risks propogating the corruption.
And hey, if you want to mix and match CPU types (uSparc 2 and 3, etc), speeds, etc, no problem either. So if you wanna upgrade your server's CPUs, there will be zero downtime, you just do it a board at a time (board = 2 or 4 CPUs).
Funny, I was just in a sales meeting with Sun in which they promised the feature of mixing CPU speeds *IN THE FUTURE*. Today you can ABSOLUTELY NOT MIX CPU TYPES and mixing clock speeds is a MARKETING PROMISE FOR THE FUTURE.
Most of your statements are lies, the remainder are simply wrong.
To take the processor offline yourself, obviously failing is not exactly the same but I thought I have had some fail without crashing. - Unless, of course you are trolling and if so, you got me.
If you were FORCED to do this because the CPU was bad, it is already TOO LATE. Perhaps you got a correctable error, e.g. on that CPU's cache and you disabled it as a precaution. But the only time I've used psradm to disable a CPU was after a PANIC AND REBOOT.
Doesn't work like that, kid. A CPU on a high-end Sun fails, and the system will keep on running. You can swap the CPU out and replace it with a new one, the system will simply pick it up, assign threads to it, and keep on running.
This is a LIE. The BEST you will get is that it detects what failed and after the PANIC AND REBOOT you will have that board or CPU disabled.
Cool, we said. No data lost, no need to shut down or even restart our app. 'Course you gotta architect your app to deal with that - like don't have just one thread that does a crucial task, 'cos there's a chance that might be on the CPU that fails. But still, it's no big deal.
Even if what you were saying is true, which it is not, what is the likelyhood that a multi *threaded* application could handle the death of an individual thread at a random instant? Probably about 0% of multithreaded applications can handle that. If they're using fork() they will probably be OK. But this is irrelevant because when a CPU fails your ENTIRE SYSTEM PANICS AND REBOOTS no matter what kind of fancy Sun hardware you are using.
The high-end Sun machines are designed for high availability. Not only will a CPU failure not crash the machine, the CPUs are hot swappable so you can replace a failed CPU without so much as a reboot.
This is a LIE. In a domainable system, a failed CPU will cause only that domain to crash. But that is it. In a domainable system, a separate domain might as well be a separate system. A failed CPU will CRASH THE KERNEL THAT KNOWS ABOUT THAT CPU!
How did a lie like this get modded to +5 when my truthful statements got modded to -1. Is this based on Sun marketing lies?
I would say that Fedora has done the graphical install wrong, as it is done in such a way that does not guarantee it staying in synch with the text installer. So critical functionality is missing from the text installer. This means that on the rare occasion that circumstances force you to do a text install, you may find that you can not properly configure the system.
Maintainability and TCO issues are highly relevant and yet never reviewed. What is the distro's history of breaking stuff with patches. Do they take care to fix more bugs than they add? If I care about the machine or I'm maintaining more than one, this can be the primary concern. I can't remember the last time a Red Hat patch broke anythins.
Are stock, stable versions of packages distributed? Or does the distribution include beta versions of software, patches to add features (and bugs), and other nonstandard modifications? In the case of Redhat, this harms reliability/stability, but you gain cool features.
Your bank, on the other hand, can't just "lose" a few transactions here and there.
Therefore the bank also needs redundancy. A 1.5 million dollar Sun machine with 64 CPUs has a lot more failure points than a single CPU machine. Any of those 64 CPUs fails, and your system will crash. That expensive Sun hardware is not much more reliable than the cheap junk. (as long as it's not -actual- junk with non-ECC memory, etc)
This is how it should be, since knowing the size of Google's hardware capacity is a very, very strategic bit of information, and the kind of thing that would allow Yahoo/MSN/whoever to get a feel for how much capital would be necessary to duplicate or improve upon it.
What you say would make sense, except that there must be thousands of people who know the size of Google's hardware capacity, or who have access to the information. You can't convince me that someone with the resources of Microsoft could not easily determine how many machines Google has.
Enterprise users do not want to compile anything. A Redhat install can be done in under 10 minutes on a fast machine with a fast network. An install that takes two days and requires manual work at every step is simply not reasonable.
Enterprise users do not generally care about performance to the extent that a different compiler option tailored for their CPU will benefit.
Enterprise users do care about the software being tested with the exact same compiler and compiler options and libraries that they are using.
Gentoo will never have widespread enterprise use. The idea is just silly.
Nonsense. It is impossible to pass a file handle from one process to another unless the 'other' process is a child. In which case you're not really passing it, anyway, are you. It is just continuing to exist along the path of execution.
There are certainly major areas, for example New York, where "Coke" means "Coca-Cola, A Registered Trademark Of The Coca Cola Company, or, perhaps if you only have another brand of cola, you may, at your option, inform me of the alternate default cola choice with a level of cheeriness also at your option, or simply bring it without further comment. In the event that you have Coca-Cola (A registered Trademark Of The Coca Cola Company) in a can but Pepsi or another cola from the fountain, you may inform me of this situation to determine whether I simply prefer the most cost-effective cola solution or the specific brand I mentioned"
I'd say your odds in New York of getting a Pepsi without comment when asking for a Coke are less than 20%. And no one will ever ask "What kind of Coke do you want" unless they mean diet, non-diet, or vanilla.
It is fraudulent to advertise this thing as an HDTV card without a huge warning that it is only an HDTV tuner and requires a huge external antenna, and does not accept component or DVI inputs.
Only a small portion of people live in an area where HDTV signals can be received with an antenna. Even fewer people already have the antenna.
Everyone in New York knows this. Everyone in the area knows that they have a good chance of working in most places but they don't do anything in the city.
This is not news. This is "guide to New York for tourists from Idaho".
svn+ssh requires all the Unix users that authenticate through ssh to be in the same group, and also requires some other magic to work. Maybe you got that magic wrong. When a permission problem occurs, your repository often needs to be "recovered" in order to fix it. Subversion's "svnadmin recover" can be used in these (usually rare) cases.
How can you call these cases rare? They're guaranteed to happen. Every time I do a commit as root, one of the db/log.* files ends up owned by root. Then I do a commit as myself, and instead of reporting the error, I get corruption. Fixing permissions and doing an svnadmin recover has solved it every time. But it made me really regret using beta software. I'm saddened to see that this has not been fixed in 1.0.
You can't use subversion as two different users. Some files in the repository end up owned by a different user, and then when you do a commit it corrupts the repository instead of reporting the error.
I recall reading that the original Unix crypt(3) algorithm was based on the Enigma machine.
Your wrongness is astounding. The fact that you were moderated up to 4 is proof that the moderation system has finally failed. I'll never read the comments on slashdot again. But I will fix this final error.
crypt(1), the file 'encryption' utility is based on a simplified Enigma. crypt(3), the password hash with an 8 character limit that can be run 100,000 times a second on a modern machine, always used DES. Modern Unixes use an md5 or blowfish based hash that doesn't have an arbitrary length limit and is designed to be much slower. (and has variable round count and a larger salt space).
If this was a concern, why wouldn't they fly a few miles out, turn around, and break the soundbarrier over international waters? NY and LA are both next to the ocean.
And you're using semantics to back up a losing argument. I said, "essentially free". But to play your game, take a look around, speech ain't exactly free lately either.
I think that your problem (besides the obvious lack of an ability to debate in a rational manner) is that you are confused about what the word "free" means in this context. Thus, "free as in speech" is to explain that "free, in this context, means as it does in ''free speech''".
The fact that you have such trouble with English could be seen as a detriment. But you might simply more familiar with another language. This list has "free software" translated into many languages. Some of these languages use a different word for "free as in freedom" vs "free as in without monetary cost".
Slashdot Mirror
This is NOT TRUE. Solaris will CRASH.
Your application will notice, because suddenly a couple of its threads will quit, but that's ok, software like Oracle already knows how to deal with failed transations.
If Oracle or Sybase or any other application I've EVER SEEN has one of its thread crash the WHOLE THING IS FUCKED. What if that thread held a lock? What you say is hopelessly stupid.
And if you can schedule a CPU board removal/swap, then there won't be ANY problems at all, as the OS will migrate threads to other CPUs and allow the removal or hardware.
There's a big difference between disabling the hardware and then removing it, and having the hardware fail. Once it has failed there has ALREADY BEEN MEMORY CORRUPTION and the system *MUST* PANIC AND REBOOT or it only risks propogating the corruption.
And hey, if you want to mix and match CPU types (uSparc 2 and 3, etc), speeds, etc, no problem either. So if you wanna upgrade your server's CPUs, there will be zero downtime, you just do it a board at a time (board = 2 or 4 CPUs).
Funny, I was just in a sales meeting with Sun in which they promised the feature of mixing CPU speeds *IN THE FUTURE*. Today you can ABSOLUTELY NOT MIX CPU TYPES and mixing clock speeds is a MARKETING PROMISE FOR THE FUTURE.
Most of your statements are lies, the remainder are simply wrong.
If you were FORCED to do this because the CPU was bad, it is already TOO LATE. Perhaps you got a correctable error, e.g. on that CPU's cache and you disabled it as a precaution. But the only time I've used psradm to disable a CPU was after a PANIC AND REBOOT.
This is a LIE. The BEST you will get is that it detects what failed and after the PANIC AND REBOOT you will have that board or CPU disabled.
Cool, we said. No data lost, no need to shut down or even restart our app. 'Course you gotta architect your app to deal with that - like don't have just one thread that does a crucial task, 'cos there's a chance that might be on the CPU that fails. But still, it's no big deal.
Even if what you were saying is true, which it is not, what is the likelyhood that a multi *threaded* application could handle the death of an individual thread at a random instant? Probably about 0% of multithreaded applications can handle that. If they're using fork() they will probably be OK. But this is irrelevant because when a CPU fails your ENTIRE SYSTEM PANICS AND REBOOTS no matter what kind of fancy Sun hardware you are using.
WHY ARE YOU LYING?
This is a LIE. In a domainable system, a failed CPU will cause only that domain to crash. But that is it. In a domainable system, a separate domain might as well be a separate system. A failed CPU will CRASH THE KERNEL THAT KNOWS ABOUT THAT CPU!
How did a lie like this get modded to +5 when my truthful statements got modded to -1. Is this based on Sun marketing lies?
I would say that Fedora has done the graphical install wrong, as it is done in such a way that does not guarantee it staying in synch with the text installer. So critical functionality is missing from the text installer. This means that on the rare occasion that circumstances force you to do a text install, you may find that you can not properly configure the system.
Maintainability and TCO issues are highly relevant and yet never reviewed. What is the distro's history of breaking stuff with patches. Do they take care to fix more bugs than they add? If I care about the machine or I'm maintaining more than one, this can be the primary concern. I can't remember the last time a Red Hat patch broke anythins.
Are stock, stable versions of packages
distributed? Or does the distribution include beta versions of software, patches to add features (and bugs), and other nonstandard modifications? In the case of Redhat, this harms reliability/stability, but you gain cool features.
Therefore the bank also needs redundancy. A 1.5 million dollar Sun machine with 64 CPUs has a lot more failure points than a single CPU machine. Any of those 64 CPUs fails, and your system will crash. That expensive Sun hardware is not much more reliable than the cheap junk. (as long as it's not -actual- junk with non-ECC memory, etc)
What you say would make sense, except that there must be thousands of people who know the size of Google's hardware capacity, or who have access to the information. You can't convince me that someone with the resources of Microsoft could not easily determine how many machines Google has.
How do you reconcile this belief with http://www.google.com/search?hl=en&lr=&ie=UTF-8&oe =UTF-8&safe=off&c2coff=1&q=%22committed+to+helping +ohio+deliver+its+electoral%22&btnG=Search
?
Enterprise users do not want to compile anything. A Redhat install can be done in under 10 minutes on a fast machine with a fast network. An install that takes two days and requires manual work at every step is simply not reasonable.
Enterprise users do not generally care about performance to the extent that a different compiler option tailored for their CPU will benefit.
Enterprise users do care about the software being tested with the exact same compiler and compiler options and libraries that they are using.
Gentoo will never have widespread enterprise use. The idea is just silly.
Nonsense. It is impossible to pass a file handle from one process to another unless the 'other' process is a child. In which case you're not really passing it, anyway, are you. It is just continuing to exist along the path of execution.
would imply that this is duplication and distribution of copyrighted material without the knowledge, consent, or license of the copyright holder.
What does this imply about the intersection of copyright law and natural rights?
No commonly used software RAID does this.
There are certainly major areas, for example New York, where "Coke" means "Coca-Cola, A Registered Trademark Of The Coca Cola Company, or, perhaps if you only have another brand of cola, you may, at your option, inform me of the alternate default cola choice with a level of cheeriness also at your option, or simply bring it without further comment. In the event that you have Coca-Cola (A registered Trademark Of The Coca Cola Company) in a can but Pepsi or another cola from the fountain, you may inform me of this situation to determine whether I simply prefer the most cost-effective cola solution or the specific brand I mentioned"
I'd say your odds in New York of getting a Pepsi without comment when asking for a Coke are less than 20%. And no one will ever ask "What kind of Coke do you want" unless they mean diet, non-diet, or vanilla.
Isn't it awfully irresponsible to have a release that did not have a corresponding release candidate?
How can anything be simpler to patch than rpm -Fvh /net/fileserver/redhat/updates/RHEL-3.0/*.rpm
It is fraudulent to advertise this thing as an HDTV card without a huge warning that it is only an HDTV tuner and requires a huge external antenna, and does not accept component or DVI inputs.
Only a small portion of people live in an area where HDTV signals can be received with an antenna. Even fewer people already have the antenna.
Why don't they just give us component inputs?
Everyone in New York knows this. Everyone in the area knows that they have a good chance of working in most places but they don't do anything in the city.
This is not news. This is "guide to New York for tourists from Idaho".
How can you call these cases rare? They're guaranteed to happen. Every time I do a commit as root, one of the db/log.* files ends up owned by root. Then I do a commit as myself, and instead of reporting the error, I get corruption. Fixing permissions and doing an svnadmin recover has solved it every time. But it made me really regret using beta software. I'm saddened to see that this has not been fixed in 1.0.
You can't use subversion as two different users. Some files in the repository end up owned by a different user, and then when you do a commit it corrupts the repository instead of reporting the error.
Your wrongness is astounding. The fact that you were moderated up to 4 is proof that the moderation system has finally failed. I'll never read the comments on slashdot again. But I will fix this final error.
crypt(1), the file 'encryption' utility is based on a simplified Enigma. crypt(3), the password hash with an 8 character limit that can be run 100,000 times a second on a modern machine, always used DES. Modern Unixes use an md5 or blowfish based hash that doesn't have an arbitrary length limit and is designed to be much slower. (and has variable round count and a larger salt space).
If you are good, and will be able to prove you are highly qualified in the interview, you can tell the truth. If you're only so-so, lie.
If this was a concern, why wouldn't they fly a few miles out, turn around, and break the soundbarrier over international waters? NY and LA are both next to the ocean.
I think that your problem (besides the obvious lack of an ability to debate in a rational manner) is that you are confused about what the word "free" means in this context. Thus, "free as in speech" is to explain that "free, in this context, means as it does in ''free speech''".
The fact that you have such trouble with English could be seen as a detriment. But you might simply more familiar with another language. This list has "free software" translated into many languages. Some of these languages use a different word for "free as in freedom" vs "free as in without monetary cost".
All these applications require Windows. Is there a way to synch my cellphone with my computer if I'm not using Windows?