Careful, there's a gotcha with IDE RAID...
on
IDE RAID Examined
·
· Score: 3, Insightful
...so be alert.
Each IDE controller can support up to two drives, a master and a slave. What happens if you hang two drives off one controller, and the "master" drive dies?
If it dies badly enough, the "slave" drive can go offline. Now you've got TWO drives in your array that aren't talking. There goes your redundancy.
If your purpose in using RAID is to have a system that can continue operating after a single drive failure, then you better think again before you hang two drives off any one controller.
As it points out in the Linux software RAID docs, you should only have one drive per IDE controller if you're really concerned about uptime. That would imply that "4 channel" RAID cards should only be used with a maximum of two drives, both set to "master", and no "slaves".
Note that this does not apply to SATA drives, as there isn't really a master-slave relationship with SATA -- all drives have separate cables and controller circuits. SATA drives are enumerated the same way as older drives for backwards compatibility with drivers and other software, but they are otherwise independent. (At least that's what I hear, I haven't actually seen one of these beasts yet...)
And of course none of this touches on controller failures, which is another issue. But if you are worried about losing drives and still staying up, then better take this into consideration when you design your dream storage system.
(I don't know about you guys, but I have lost several drives over the years, and not one controller...)
...and a silly one. Like people would be willing to be locked into a pressurized metal tube and hurled at hundreds of miles per hour....oh wait, we do that now, but we call them airplanes, and they're much more useful because you can demolish old buildings with them.
Never mind.
One can think of a zillion reasons why this would never work. For one thing, you'd have to build a pair of tubes between each possible pair of destinations -- that's a lot of combinations....no wait, even airplanes don't work that way any more. Now they mainly fly to and from a few centralized hubs.
Never mind.
And of course, there's the tremendous expense of boring zillions of miles of tunnels, not to mention the exorbitant cost of obtaining the necessary underground right-of-way in a property-crazed world....oh wait, soon we'll have nanotechnology and sentient machines and everything will be free except food and land, and the only people who have money will be the ones who are leasing their underground rights to the tube train companies / mining conglomerates.
Never mind.
* * *
Gerry O'Neill wrote extensively on this idea in his book '2081'. (Gerry also did pioneering work on space colony concepts, solar power satellites, and satellite phones [Geostar], among many others. He also invented the electromagnetic mass driver and the colliding beam synchrotron, to name just a few. The guy got around.)
Long ago I was compelled to the conclusion that I was the only person who owned a copy of this book, or had even read it. Then, purely by accident, I discovered someone else like this, who happened to be a close friend. So we compared notes. She thought the ideas in the book would never happen, because they would be just too difficult to build. I disagreed, saying that most of the necessary technology already existed (it does, and did). Then I went on to say that I found the book rather silly because it describes life in 2081 as being so similar to 1981, with people having jobs and property and such. Now whenever we see each other, we look at each other as if we both had guacamole coming out of our ears.
Once I wrote a really cool science fiction story about a future when the Earth's crust is honeycombed with tubes of all sorts, evacuated and otherwise, that are mainly used to move people and packages around quickly, cheaply and safely. All the debris from the (fully automated) tunneling operations was used to build levees around the continents to hold back the constantly rising sea level (except for the places where poor people lived, they had to eke out a living on vast lashed-together fleets of rotting derelict fishing boats that were now useless because all the fish were extinct). Then I spell-checked it, and Word changed all the words to 'Mine!' and then the machine blue-screened and the hard disk vaporized.
There are numerous researchers who have said that the multiverse (of which our universe is of course just a tiny part) has a total information content of essentially zero.
Saw Episode 1 in two different digital theaters. Had a chance to talk with some of the engineers who developed these technologies.
The first theater used an infrared-modulated liquid-crystal image light valve, developed by Hughes. An interesting feature of this technique is that the resolution is continuously adjustable (there are no pixels in the projector). Because the transfer function of the liquid crystal is nonlinear, it needs to be calibrated frequently, and that process (at that time) was not very automated -- perhaps that is why the TI mirror chip dominates this application today.
The second theater used the TI mirror chip. It provides fixed HDTV resolution, although it is possible to run this chip at various frame rates.
Both theaters used identical sources, at the same frame rate and HDTV resolution. (The projector uses an appropriate lens to stretch the image into the desired aspect ratio.) The compressed and encrypted movie was stored on a redundant array of disk drives (RAID5 I assume) that had a total capacity of 350 GB. An industry standard [don't remember the name] box was used to decrypt and uncompress the frames on the fly. Episode 1 used up about 250 GB, compressed.
Obviously, the compression used in this application is nowhere near as severe as the compression used on DVDs. I suspect this is one of the most important reasons why the image quality is so remarkably good. Yes, this is merely an HDTV picture writ large, but you will never see an image this good on any home theater -- at least not until you can get a 250 GB file at home....:-)
The biggest improvement I noticed with (both of these) digital projections, compared with film, was the total lack of frame jitter. With film, there is always some wear and tear of the sprocket holes, and even a brand-new print will have the image jittering back and forth a little bit. (Imax, with its film registration pins, is largely immune to this problem.) Digital projection eliminates this annoyance forever.
Although the images looked just fabulous, everything was not perfect. The end credits, in particular, showed some artifacts, although these were not really obvious unless you walked up toward the front of the theater. Perhaps these were due to choices made during the compression process -- after all, Episode 1 was not a fully digital movie, and in fact this version was digitized from a completed film print. It will be interesting to see how a fully digital Episode 2 compares in this regard.
Ultimately, I would like to see digital theaters break away from their current limitations of resolution and (especially) frame rate, but I expect nobody will invest in these kinds of improvements until digital projection is much more widely accepted.
Slashdot Mirror
...so be alert.
Each IDE controller can support up to two drives, a master and a slave. What happens if you hang two drives off one controller, and the "master" drive dies?
If it dies badly enough, the "slave" drive can go offline. Now you've got TWO drives in your array that aren't talking. There goes your redundancy.
If your purpose in using RAID is to have a system that can continue operating after a single drive failure, then you better think again before you hang two drives off any one controller.
As it points out in the Linux software RAID docs, you should only have one drive per IDE controller if you're really concerned about uptime. That would imply that "4 channel" RAID cards should only be used with a maximum of two drives, both set to "master", and no "slaves".
Note that this does not apply to SATA drives, as there isn't really a master-slave relationship with SATA -- all drives have separate cables and controller circuits. SATA drives are enumerated the same way as older drives for backwards compatibility with drivers and other software, but they are otherwise independent. (At least that's what I hear, I haven't actually seen one of these beasts yet...)
And of course none of this touches on controller failures, which is another issue. But if you are worried about losing drives and still staying up, then better take this into consideration when you design your dream storage system.
(I don't know about you guys, but I have lost several drives over the years, and not one controller...)
...and a silly one. Like people would be willing to be locked into a pressurized metal tube and hurled at hundreds of miles per hour. ...oh wait, we do that now, but we call them airplanes, and they're much more useful because you can demolish old buildings with them.
...no wait, even airplanes don't work that way any more. Now they mainly fly to and from a few centralized hubs.
...oh wait, soon we'll have nanotechnology and sentient machines and everything will be free except food and land, and the only people who have money will be the ones who are leasing their underground rights to the tube train companies / mining conglomerates.
Never mind.
One can think of a zillion reasons why this would never work. For one thing, you'd have to build a pair of tubes between each possible pair of destinations -- that's a lot of combinations.
Never mind.
And of course, there's the tremendous expense of boring zillions of miles of tunnels, not to mention the exorbitant cost of obtaining the necessary underground right-of-way in a property-crazed world.
Never mind.
* * *
Gerry O'Neill wrote extensively on this idea in his book '2081'. (Gerry also did pioneering work on space colony concepts, solar power satellites, and satellite phones [Geostar], among many others. He also invented the electromagnetic mass driver and the colliding beam synchrotron, to name just a few. The guy got around.)
Long ago I was compelled to the conclusion that I was the only person who owned a copy of this book, or had even read it. Then, purely by accident, I discovered someone else like this, who happened to be a close friend. So we compared notes. She thought the ideas in the book would never happen, because they would be just too difficult to build. I disagreed, saying that most of the necessary technology already existed (it does, and did). Then I went on to say that I found the book rather silly because it describes life in 2081 as being so similar to 1981, with people having jobs and property and such. Now whenever we see each other, we look at each other as if we both had guacamole coming out of our ears.
Once I wrote a really cool science fiction story about a future when the Earth's crust is honeycombed with tubes of all sorts, evacuated and otherwise, that are mainly used to move people and packages around quickly, cheaply and safely. All the debris from the (fully automated) tunneling operations was used to build levees around the continents to hold back the constantly rising sea level (except for the places where poor people lived, they had to eke out a living on vast lashed-together fleets of rotting derelict fishing boats that were now useless because all the fish were extinct). Then I spell-checked it, and Word changed all the words to 'Mine!' and then the machine blue-screened and the hard disk vaporized.
There are numerous researchers who have said that the multiverse (of which our universe is of course just a tiny part) has a total information content of essentially zero.
Those of you with too much time on your hands may enjoy Schmidhuber's 1996 paper, A Computer Scientist's View of Life, the Universe, and Everything.
And check out the Everything List archives too.
Saw Episode 1 in two different digital theaters. Had a chance to talk with some of the engineers who developed these technologies.
:-)
The first theater used an infrared-modulated liquid-crystal image light valve, developed by Hughes. An interesting feature of this technique is that the resolution is continuously adjustable (there are no pixels in the projector). Because the transfer function of the liquid crystal is nonlinear, it needs to be calibrated frequently, and that process (at that time) was not very automated -- perhaps that is why the TI mirror chip dominates this application today.
The second theater used the TI mirror chip. It provides fixed HDTV resolution, although it is possible to run this chip at various frame rates.
Both theaters used identical sources, at the same frame rate and HDTV resolution. (The projector uses an appropriate lens to stretch the image into the desired aspect ratio.) The compressed and encrypted movie was stored on a redundant array of disk drives (RAID5 I assume) that had a total capacity of 350 GB. An industry standard [don't remember the name] box was used to decrypt and uncompress the frames on the fly. Episode 1 used up about 250 GB, compressed.
Obviously, the compression used in this application is nowhere near as severe as the compression used on DVDs. I suspect this is one of the most important reasons why the image quality is so remarkably good. Yes, this is merely an HDTV picture writ large, but you will never see an image this good on any home theater -- at least not until you can get a 250 GB file at home....
The biggest improvement I noticed with (both of these) digital projections, compared with film, was the total lack of frame jitter. With film, there is always some wear and tear of the sprocket holes, and even a brand-new print will have the image jittering back and forth a little bit. (Imax, with its film registration pins, is largely immune to this problem.) Digital projection eliminates this annoyance forever.
Although the images looked just fabulous, everything was not perfect. The end credits, in particular, showed some artifacts, although these were not really obvious unless you walked up toward the front of the theater. Perhaps these were due to choices made during the compression process -- after all, Episode 1 was not a fully digital movie, and in fact this version was digitized from a completed film print. It will be interesting to see how a fully digital Episode 2 compares in this regard.
Ultimately, I would like to see digital theaters break away from their current limitations of resolution and (especially) frame rate, but I expect nobody will invest in these kinds of improvements until digital projection is much more widely accepted.