I'd also point out that with most standard RAID scenarios, when a disk fails the rest of the disks in that RAID set generally become immensely more busy than their normal workload further note that with parity based raid you get a double whammy of extra load. not only does the rebuild cause extra load but any read ops that would normally hit the failed drive require a read of ALL drives to satisfy.
and the rebuild operation itself needs to read ALL data off ALL drives.
I would like to put forward the notion that perhaps they do it because they have a (perhaps quite justified) fear that americans calling in will assume that anyone with an accent is sub-standard and cannot help them, therefore will refuse it and make a fuss from the get go. from the customers side if i get an accent that doesn't match with who i'm calling it generally means one or more of the following.
1: the buisness is being a cheapskate, this doesn't bode well for the rest of the support request, this is especially true if the line quality is poor and/or the person doesn't just have a foriegn accent but also seems to have a poor command of english. 2: the buisness is probablly outsourcing, this means that not only will the drone be unlikely to know (that applies to drones anywhere though) but the drones entire chain of command probablly also don't really know.
However, when that failure point is reached is at a random point within the distribution so while the probability of another failure at any point in time is not zero it is pretty small.
There are three real dangers with raid
The first is that arrays are typically built out of identical drives, usually drives from the same batch and then all the drives are run for the same time periods. This means that if there is a design or manufacturing fault that causes a failure peak at a certain number of operational hours there is a good chance that more than one drive in your array will fail at about the same time.
The second is that the drives in an array are typically in one machine, running off one power supply (or one pair of redundant power supplies) and connected to one controller. This means that faults with other hardware in the machine can destroy multiple hard drives at once.
The third is failure of the controller. In many cases the controller stores information on how the data is set up within its own non-volatile memory (some better controllers do store it on the disks themselves) while this doesn't destroy the actual data it can easilly put it beyond the ability of non-experts to reassemble the array in a way that gets the data back (and if they make a mistake they can easilly destory the data they were trying to recover). There is also the problem that getting a suitable replacement controller may be difficult.
Remember though that the cost of a hard drive failure can be considerablly more than the cost of the replacement drive.
first you have the cost of the actual replacement, easy enough if you can do it yourself but if you have to pay someone to do it for you....
then you have the value of the time spent getting your system back into a usable state, if you have a raid setup this is small but if you are relying on data only backups or you have problems with your full backup it could take MUCH longer.
then you have the value to you of any data that is lost forever.
If anything, RAID should make your hard disk access a lot faster. That is, unless you go for software RAID, which will put a hit on your processor. afaict Linux software raid is actually pretty good nowadays at least as long as you stick to the basic raid levels
beware of the very common fake hardware (e.g. really software but with some bios and driver magic to make the array bootable and generally behave like hardware raid from the users point of view) controllers. Theese often have far worse performance than linux software raid and many of them only support windows.
iirc on a transfer you have to buy an extra year but that gets added to your domains existing expiry date not to the date you made the transfer so you don't really lose anything other than a negligable ammount of interest on the money by transfering sooner rather than later.
i've learnt the hard way that with troublesome registrars a transfer just before expiry is a bad idea for both your sanity and your chances of keeping the domain, make the transfers while you still have plenty of time left to deal with screwups.
MSIs are one of the best ideas for Windows in a while note that MSIs are nothing new, office was using that system (with a stub exe to make sure the microsoft installer was installed first) from office 2000 onwards and i think win2k shipped with it as standard.
Some people seem to think that the fact that Debian does things differently from Mandriva that does it different than Fedora is what makes the distribution "special". Be that as it may, I think it's only hurting Linux users as a whole. It is certainly a problem, trouble is the moment you get standards bodies involved progress slows to a glacial pace, distros understandablly want to be able to improve the way thier libraries are packaged to meet thier users needs.
In many ways windows represents the opposite extreme, they are only just now (with the 64 bit versions) dropping support for win16 apps, windows contains a huge ammount of cruft for backwards compatibility.
another problem is that library authors act in ways that make it very hard to build binaries on a newer system that will run correctly on an older system.
a final problem is that menu systems vary so even once you get a binary on the system in a way it will run (though i belive freedesktop.org is trying to work on getting this standardised)
tools like autopackage try to work arround theese problems but doing so is easier said than done.
some modern petrol cars are already quiet enough that you can't hear them if there is a bit of background noise (think a side road off a busy main road).
true, i was just proving the point that XOR and AND together were enough, not looking for the optimal ways of implementing stuff.
in electronics we don't tend to use XOR as a building block because XOR is a relatively expensive function, the primitives that you get from transistor cuircuits are NAND, NOR and NOT (a NOT gate is essentially a 1 input NAND or NOR gate).
didn't the enron guys only get in real (criminal) trouble after the buisness had already collapsed because they could no longer keep up the illusion of profitability?
On the other hand a system administrator might force the system onto a wrong time, but then, he is a sysadmin, so must know better. rofl
yes if you have a locked down environment where a sysadmin who understands timezones has exclusive control over the clocks on all the machines its not a problem (on either windows or *nix). I'm willing to bet the majority of systems do not fall into that category.
or you use an os which doesn't have the new timezone settings, the user notices the clock is "wrong" and "corrects" it.
The result of this is that the systems idea of UTC ends up an hour out of step with reality.
most non-programmers really don't get the subtulties of the situation and the os provides no luser friendly interface to fix the issue, at least thats what its like on windows i dunno about the modern linux desktops.
Correct me if i'm wrong but a computer cannot be created from an AND gate and an XOR gate you are wrong, XOR together with AND is enough (neither is sufficiant on its own)
NOT A = 1 XOR A A NAND B = NOT (A AND B)= 1 XOR (A AND B) A NOR B = (NOT A) AND (NOT B) = (1 XOR A) AND (1 XOR B) A OR B = NOT ((NOT A) AND (NOT B)) = 1 XOR ((1 XOR A) AND (1 XOR B))
Lets say PJ is a crack whore who gives blowjobs for drugs. Let's say SCO introduces this evidence in court. How does this help their case? you are making the mistake that all sco cares about is what the court decides.
a company whose executives are trying to execute the dump part of a pump and dump really doesn't want thier name to be dragged through the dirt. So the pretty clear aim is to try and silence or discredit the person who is doing it to them.
they are scared because they know because to most people 128k skip free mp3 is as good as CD (yes there are a few with especially good hearing or who have trained themselves to do it who can hear the difference) and can be easilly burnt to a CD or copied to a portable player with full track seperation maintained.
also after the initial loss of making the mp3 there is no further loss no matter how many times it is redistributed.
compare this to copying a LP or CD to tape for friends, there you have a lot of generation loss and on no copys do you get easy and quick track selection (yes track selection is pretty easy on a LP)
the music industry want some copying (even though they can't/won't admit it) but they don't want high quality copies moving easilly round the world with no generation loss.
CD was an attempt to make the dominant digital audio format a read-only one but CD-R, MP3 and flash based portable audio players screwed up that plan.
obviously there needs to be a chip at each end to compress the signal to/from mpeg2 that'd sort out the bandwith! start watching something on analog TV, then switch to the same station on digital TV, notice about half a second is repeated.
MPEG can only acheive high compresion by introducing latency that would be unacceptable for interactive use.
Indeed, let's also include the graphics card with the monitor instead of the computer and run an X server on the monitor and connect it through ethernet. If we in addition connect the keyboard and mouse directly to that monitor, we could even put it remote from the actual computer if we wish to. We just need a nice name for that monitor/keyboard/mouse combination running X. Well, what about calling it "X Terminal"? [wikipedia.org] and lets do that while running 3D games or playing videos.
1024x768*24*60=1,132,462,080bits/sec
and thats a pretty low end display
the fact is that neither firewire 800 or gigabit ethernet is enough to support a new frame every refresh for even a fairly low end display nowadays (unless you use compression but that causes nasty latency).
10 gigabit ethernet would do it, so would an external variant of PCI-EXPRESS (i've seen pictures of a professional video processing box that brought PCI-EXPRESS out to an external cable). but both of theese require connectors that are even more exotic than those currently used for digital video.
But what is more interesting is that the link IS NOT A DIRECT LINK. Instead it redirects through DoubleClick for some reason. I am not trying to make this sound sinister, but I found that a little odd. i don't, most likely it just means OSTG want to charge intel for clicks on it.
I hate those adapter things, they always feel like they are going to snap off those pins. hmm, they've always seemed pretty firm to me, in fact very similar to the old 25-9 way serial adaptors and DVI has never seemed like a damage prone connector standard to me either (its very similar to the D connectors and only once have i seen a pin bent on one of those)
Slashdot Mirror
I'd also point out that with most standard RAID scenarios, when a disk fails the rest of the disks in that RAID set generally become immensely more busy than their normal workload
further note that with parity based raid you get a double whammy of extra load. not only does the rebuild cause extra load but any read ops that would normally hit the failed drive require a read of ALL drives to satisfy.
and the rebuild operation itself needs to read ALL data off ALL drives.
I would like to put forward the notion that perhaps they do it because they have a (perhaps quite justified) fear that americans calling in will assume that anyone with an accent is sub-standard and cannot help them, therefore will refuse it and make a fuss from the get go.
from the customers side if i get an accent that doesn't match with who i'm calling it generally means one or more of the following.
1: the buisness is being a cheapskate, this doesn't bode well for the rest of the support request, this is especially true if the line quality is poor and/or the person doesn't just have a foriegn accent but also seems to have a poor command of english.
2: the buisness is probablly outsourcing, this means that not only will the drone be unlikely to know (that applies to drones anywhere though) but the drones entire chain of command probablly also don't really know.
However, when that failure point is reached is at a random point within the distribution so while the probability of another failure at any point in time is not zero it is pretty small.
There are three real dangers with raid
The first is that arrays are typically built out of identical drives, usually drives from the same batch and then all the drives are run for the same time periods. This means that if there is a design or manufacturing fault that causes a failure peak at a certain number of operational hours there is a good chance that more than one drive in your array will fail at about the same time.
The second is that the drives in an array are typically in one machine, running off one power supply (or one pair of redundant power supplies) and connected to one controller. This means that faults with other hardware in the machine can destroy multiple hard drives at once.
The third is failure of the controller. In many cases the controller stores information on how the data is set up within its own non-volatile memory (some better controllers do store it on the disks themselves) while this doesn't destroy the actual data it can easilly put it beyond the ability of non-experts to reassemble the array in a way that gets the data back (and if they make a mistake they can easilly destory the data they were trying to recover). There is also the problem that getting a suitable replacement controller may be difficult.
Remember though that the cost of a hard drive failure can be considerablly more than the cost of the replacement drive.
first you have the cost of the actual replacement, easy enough if you can do it yourself but if you have to pay someone to do it for you....
then you have the value of the time spent getting your system back into a usable state, if you have a raid setup this is small but if you are relying on data only backups or you have problems with your full backup it could take MUCH longer.
then you have the value to you of any data that is lost forever.
If anything, RAID should make your hard disk access a lot faster. That is, unless you go for software RAID, which will put a hit on your processor.
afaict Linux software raid is actually pretty good nowadays at least as long as you stick to the basic raid levels
beware of the very common fake hardware (e.g. really software but with some bios and driver magic to make the array bootable and generally behave like hardware raid from the users point of view) controllers. Theese often have far worse performance than linux software raid and many of them only support windows.
iirc on a transfer you have to buy an extra year but that gets added to your domains existing expiry date not to the date you made the transfer so you don't really lose anything other than a negligable ammount of interest on the money by transfering sooner rather than later.
i've learnt the hard way that with troublesome registrars a transfer just before expiry is a bad idea for both your sanity and your chances of keeping the domain, make the transfers while you still have plenty of time left to deal with screwups.
MSIs are one of the best ideas for Windows in a while
note that MSIs are nothing new, office was using that system (with a stub exe to make sure the microsoft installer was installed first) from office 2000 onwards and i think win2k shipped with it as standard.
Some people seem to think that the fact that Debian does things differently from Mandriva that does it different than Fedora is what makes the distribution "special". Be that as it may, I think it's only hurting Linux users as a whole.
It is certainly a problem, trouble is the moment you get standards bodies involved progress slows to a glacial pace, distros understandablly want to be able to improve the way thier libraries are packaged to meet thier users needs.
In many ways windows represents the opposite extreme, they are only just now (with the 64 bit versions) dropping support for win16 apps, windows contains a huge ammount of cruft for backwards compatibility.
another problem is that library authors act in ways that make it very hard to build binaries on a newer system that will run correctly on an older system.
a final problem is that menu systems vary so even once you get a binary on the system in a way it will run (though i belive freedesktop.org is trying to work on getting this standardised)
tools like autopackage try to work arround theese problems but doing so is easier said than done.
because thats an old budget.
wikipedias costs are skyrocketing along with its size, the only way they can keep up with demand is to buy ever more server hardware.
some modern petrol cars are already quiet enough that you can't hear them if there is a bit of background noise (think a side road off a busy main road).
true, i was just proving the point that XOR and AND together were enough, not looking for the optimal ways of implementing stuff.
in electronics we don't tend to use XOR as a building block because XOR is a relatively expensive function, the primitives that you get from transistor cuircuits are NAND, NOR and NOT (a NOT gate is essentially a 1 input NAND or NOR gate).
didn't the enron guys only get in real (criminal) trouble after the buisness had already collapsed because they could no longer keep up the illusion of profitability?
the opensource initiatives open source definition and stallmans free software definition are pretty much the same.
On the other hand a system administrator might force the system onto a wrong time, but then, he is a sysadmin, so must know better.
rofl
yes if you have a locked down environment where a sysadmin who understands timezones has exclusive control over the clocks on all the machines its not a problem (on either windows or *nix). I'm willing to bet the majority of systems do not fall into that category.
or you use an os which doesn't have the new timezone settings, the user notices the clock is "wrong" and "corrects" it.
The result of this is that the systems idea of UTC ends up an hour out of step with reality.
most non-programmers really don't get the subtulties of the situation and the os provides no luser friendly interface to fix the issue, at least thats what its like on windows i dunno about the modern linux desktops.
This means that people who change the underlying clock to make the screen numbers look right will be causing more damage than good
right you are
this issue is also seen sometimes due to people who set the local time right but never bother with the timezone setting.
and its very hard to explain all this to lusers who don't userstand timezones
Correct me if i'm wrong but a computer cannot be created from an AND gate and an XOR gate
you are wrong, XOR together with AND is enough (neither is sufficiant on its own)
NOT A = 1 XOR A
A NAND B = NOT (A AND B)= 1 XOR (A AND B)
A NOR B = (NOT A) AND (NOT B) = (1 XOR A) AND (1 XOR B)
A OR B = NOT ((NOT A) AND (NOT B)) = 1 XOR ((1 XOR A) AND (1 XOR B))
Lets say PJ is a crack whore who gives blowjobs for drugs. Let's say SCO introduces this evidence in court. How does this help their case?
you are making the mistake that all sco cares about is what the court decides.
a company whose executives are trying to execute the dump part of a pump and dump really doesn't want thier name to be dragged through the dirt. So the pretty clear aim is to try and silence or discredit the person who is doing it to them.
with the probe in space there is little hope of finding out what wen't wrong, back on earth on the other hand.....
also it sounds like there is the off chance it got its sample so may as well find out
they are scared because they know because to most people 128k skip free mp3 is as good as CD (yes there are a few with especially good hearing or who have trained themselves to do it who can hear the difference) and can be easilly burnt to a CD or copied to a portable player with full track seperation maintained.
also after the initial loss of making the mp3 there is no further loss no matter how many times it is redistributed.
compare this to copying a LP or CD to tape for friends, there you have a lot of generation loss and on no copys do you get easy and quick track selection (yes track selection is pretty easy on a LP)
the music industry want some copying (even though they can't/won't admit it) but they don't want high quality copies moving easilly round the world with no generation loss.
CD was an attempt to make the dominant digital audio format a read-only one but CD-R, MP3 and flash based portable audio players screwed up that plan.
When will you Americans learn you cannot have it both ways.
they CAN have it both ways because they are the big bully and you aren't
sad but true
(the eu could be a big bully if they wanted to but we argue amongst ourselves too much for that)
obviously there needs to be a chip at each end to compress the signal to/from mpeg2 that'd sort out the bandwith!
start watching something on analog TV, then switch to the same station on digital TV, notice about half a second is repeated.
MPEG can only acheive high compresion by introducing latency that would be unacceptable for interactive use.
Indeed, let's also include the graphics card with the monitor instead of the computer and run an X server on the monitor and connect it through ethernet. If we in addition connect the keyboard and mouse directly to that monitor, we could even put it remote from the actual computer if we wish to. We just need a nice name for that monitor/keyboard/mouse combination running X. Well, what about calling it "X Terminal"? [wikipedia.org]
and lets do that while running 3D games or playing videos.
1024x768*24*60=1,132,462,080bits/sec
and thats a pretty low end display
the fact is that neither firewire 800 or gigabit ethernet is enough to support a new frame every refresh for even a fairly low end display nowadays (unless you use compression but that causes nasty latency).
10 gigabit ethernet would do it, so would an external variant of PCI-EXPRESS (i've seen pictures of a professional video processing box that brought PCI-EXPRESS out to an external cable). but both of theese require connectors that are even more exotic than those currently used for digital video.
But what is more interesting is that the link IS NOT A DIRECT LINK. Instead it redirects through DoubleClick for some reason. I am not trying to make this sound sinister, but I found that a little odd.
i don't, most likely it just means OSTG want to charge intel for clicks on it.
I hate those adapter things, they always feel like they are going to snap off those pins.
hmm, they've always seemed pretty firm to me, in fact very similar to the old 25-9 way serial adaptors and DVI has never seemed like a damage prone connector standard to me either (its very similar to the D connectors and only once have i seen a pin bent on one of those)
i don't belive DVI/HDMI carry any delays that are significant enough to be noticed.
digital TV is a totally different ball game, its optimised for bandwidth efficiancy rather than latency.