I don't "overlook" that at all; since that's what real-world hardware is like, real-world fault tolerant software needs to deal with it.
Not at all. No one outside of ignorant home users tolerate such hardware. In the real world, hardware that ignores cache flushes (ie. does not guarantee the ordering of writes) is absolutely not tolerated. This is not a fault, and it can't be worked around in software.
Data that doesn't ever get written will obviously be lost. However, it appears that ZFS problems with such hardware go beyond that.
Again, you are missing the point, this has nothing to do with unwritten data. Bad hardware as described above can still damage traditional filesystems beyond repair, though usually the corruption is more localized. It's not that it doesn't happen, just that you usually won't find out immediately unless it's catastrophic. By the time you discover it, the errors may have already propagated to your backups as well, and then you are screwed.
I should have known better than to post in an Apple focused thread. It is no use trying to inform the willfully ignorant.
What good is a fault tolerant file system if it isn't tolerant of faults?
This is not about faults; ZFS handles those fine. This is about hardware which behaves badly.
You should not talk so authoritatively when you are so obviously ignorant of the subject. What you are implying, is that a filesystem does not care about the ordering of writes, and that is absolutely absurd. The ordering of writes is more critical for copy on write filesystems like ZFS, but in neither case is your data safe on bad hardware.
Like you point out, there is lot of bad hardware out there. What you overlook is that existing filesystems have no facilities to catch, much less correct such errors or corruption; that is why it is called silent data corruption. Even filesystems like HFS+ only journal metadata, so this is a lot more common than you realize.
ZFS can be improved on misbehaving hardware, but that still won't fix bad hardware. Those improvements will only allow the possibility to recover to a consistent state, but data will still be lost.
This is almost entirely nonsense. I have been following the zfs-discuss list for years, and almost no one has lost data. There have been a few bugs which could in rare cases render your data inaccessible, but they almost always have workarounds, and do get fixed.
The data loss and corruption that the parent is talking about is the fault of crap hardware. In almost every case, USB is involved, or more rarely the lack of ECC ram. It is true that ZFS is less tolerant of bad hardware. Note, faulty good hardware is not considered bad; that is reserved for garbage which (for instance) lies to the OS about flushing the disk cache. With such hardware, it is impossible for any filesystem to function reliably.
USB and Firewire bridges are notorious for this. If you care about your data, you should run the other way if you happen upon one. ZFS works great on good hardware though. With directly attached disks and ECC ram, there is no cause for concern.
We should be building IFR style plants even if Polywell fusion works out in the next couple of years.
For no other reason than to have somewhere to dispose of the enormous amount of nuclear waste that has been and continues to be generated from conventional nuclear plants. Almost the entirety of that "waste" can be used as fuel in these reactors, rather than sitting in ponds where it will inevitably bite us in the ass.
The true waste from an IFR is very minimal and easily managed, the plants are passively safe, and the reprocessing is proliferation proof.
I am all for investment in fusion technology, but that money should be better spent. Even if the Tokamak approach can be made to work, it will never be economically viable. It is extremely expensive, and there is no hope of scaling it down; the physics requires such an enormous and complex machine.
In the mean time, there are a number of other very promising approaches which continue to be neglected, and these could be funded at a small fraction of the cost. What is very frustrating is that most of these have been around for a long time, and some were even cancelled so that our futile pursuit of Tokamaks could continue.
It is not that Communism didn't or doesn't work, but rather it has never existed as Marx envisioned. The reform has never succeeded. Communism, like Democracy, exists only as a fantasy.
As an ideal, it is not half bad. Sadly, the worst kind of humans always manage to find a way to ruin things for everyone, regardless of the government.
This is excellent news. As you allude, PRAM will finally make good on the promise of solid state storage. It will allow for both higher reliability and deterministic performance, without the ludicrous internal complexity of Flash based devices.
I can't help but cringe every time I hear the terms Flash and SSD used interchangeably. If anything, the limitations inherent to Flash devices described by the GP mean they have more in common with a hard disk, as they also have an inherent physical "geometry" which must be considered.
PRAM will basically look like a simple linear byte array, without all the nonsense associated with Flash. Even if Flash retains a (temporary) advantage in density, it will never compete with hard disks on value for bulk storage, nor will it ever compete with a proper SSD on a performance basis. It makes for a half-assed "SSD", and I can't wait for it to disappear.
The advantage of a breeder reactor like the IFR is that it uses a fuel cycle in which this is effectively impossible. You would need to run a completely different fuel cycle, and likewise, the type of reprocessing facilities required are completely different. This is a good thing.
If we really need more Plutonium, we should be looking to dismantle our weapons stockpile instead. It is way way beyond what could ever be considered reasonable.
As you suggest, there are now a number meltdown-proof reactor designs. These are not merely engineered with "infallible" safety mechanisms, but are fundamentally meltdown-proof by their very design. As long as the laws of physics hold, which is a reasonably safe assumption, there is no risk of meltdown.
While the Pebble bed reactor is safe though, the nature of the pebbles make for very difficult reprocessing, and otherwise still pose a long term waste management problem.
Nuclear is the clear winner for clean, environmentally friendly energy production, but I would recommend pointing people to the Integral Fast Reactor instead. An added benefit would be that such a design could also solve our current nuclear waste problems, by recycling it for use in such reactors. The true waste after recycling is both very minimal and very short lived by comparison.
Sun has been making quite a bit of noise in the storage architecture world with their use of SSDs as intermediate cache to improve reading and writing speeds.
You are conflating Sun's claims here, as the performance gains from using SSDs in their configuration are not generally applicable to other Flash based systems.
ZFS will use SSDs in two very different ways: as cache(L2ARC) devices, and log devices. The cache devices are for improving read IOPs on a mostly static working data set, and a large Flash-based SSD is fine in this scenario. The log devices are for reducing the latency of synchronous writes, and a small DRAM-based SSD is used in this case.
Apple does say that the 17" MBP has a 60% greater gamut, so perhaps it has been improved. I haven't have a chance to compare though.
The LED displays do have other nice features, and are easy on the eyes. The glossy could be easily fixed with an anti-reflective coating, and as you put it, this should not be a problem on a $3000 machine.
Agreed, they both suck. I think what you are referring to is the anti-reflective film that was often used on CRTs. This provides the advantages of both glossy and matte, with none of the disadvantages. It is strange how people seem to have forgotten all about it.
As a reminder, examine the anti-reflective coating on a good pair of glasses; it is extremely effective. If Apple would simply apply such a coating to their displays, this whole matter could be put to rest. Matte/glossy does not need to be a choice...
Just when wide gamut LCDs are approaching the range of colors once possible on CRTs, Apple has taken yet another step backwards with their new LED backlight displays.
My LED MacBook has a spectacularly bad display, so I went to visit the local Apple store to see if this was typical. Sadly it is, and what's more, it looks like all of Apple's LED displays are vastly inferior to that of my old iMac G5. (which has an S-IPS panel and conventional fluorescent backlight)
Color wise, the LED MacBook Pro and Cinema Display are better than the MacBook, but they are all shamefully bad, and definitely worthy of a "worst in the industry" rating. (at least color-wise)
In this case, powers of 2 are more convenient for both the machine and user, as that is how virtually all software and operating systems display disk usage.
It is bad form to present data in an inconvenient format for the user, and using the SI units does exactly that.
Richard Bennett could not more obviously present himself as a shill for comcast and the like. The article is complete nonsense, attempting to portray bittorrent as the enemy of the Internet.
This idiot drones on about the "ungentlemanly" conduct of using UDP for such purposes, but conveniently avoids the fact that comcast/sandvine caused this mess by injecting face TCP resets, to break bittorrent's TCP connections. Well, what does he expect would happen?
Obviously, UDP is not a good choice for bulk transfers as it lacks congestion control, but lets be fair about where the fault lies. This is not something that can be worked around at the application level, and after being pushed into this corner, there is little else that can be done to work around their abuses of the TCP protocol.
I can not recommend these books enough. Feynman does a brilliant job of bringing the concepts of physics to life.
All together, they are quite extensive, but the individual topics are brief enough to digest in one sitting. Wether you only have a passing interest in physics, or a graduate degree in the field, you will find that there is much to appreciate in these lectures.
Even for those simply taking physics as requirement, I think that these would give you a real appreciation of the field, and probably make the classes a lot easier at that.
Well, perhaps it is in part my physics background, but I didn't get that impression at all. It is a brilliant idea, and even if you aren't familiar with Dr. Bussard, the man knows what he is talking about. He was simply old, somewhat bitter, and impatient--seemingly with good cause. Sadly, he won't see the results of his endeavors, but the research is solid, and thankfully, the navy is following it up.
In any case it has nothing to do with conspiracy theories or blaming fellow scientists. The fact is, basically nothing aside from Tokamak research is funded at a significant level.
Font rendering on the Mac is great, and one of the primary reasons I stick with the platform. The article you cite is nonsense; if you examine the images, it is clear that sub-pixel rendering isn't even enabled for the Mac. (The pure greyscale font images make that plainly obvious...)
It may not be as readable with 4 point fonts, but that is a better compromise than poorly rendering all fonts.
Perhaps most importantly though, the Mac actually comes with good fonts. I don't know about the other languages, but the Japanese fonts in XP are horrendously bad.
No need to compare with 15k rpm drives; flash disks lose spectacularly to low rpm laptop drives for random write performance. For obvious reasons though, no one ever tests random write performance. Manufacturers also rarely report random write IOPS.
Flash is great, if your disk is basically read-only.
Regular fluorescents also contain mercury (a lot more of it) and they also end up in land fills. While CFLs are relatively newer, most people don't even know that conventional fluorescents contain mercury, so there is obviously a problem. Of course, the problem extends to many other hazardous wastes, such as electronics, NiCd batteries, paint, etc.
Whatever the case, there need to be more incentives to safely dispose of hazardous waste. Disposal should be easily accessible, and the costs should be included in the purchase price, with no exceptions. Someone in government needs to get to work on this, because the problem won't fix itself. As is the case with such externalities, corporations are happy to overlook them to our detriment.
Unless they have found a way to record 100 layers at once, it will take nearly forever to record a disc with this new format. For the same reason, the proposed 3+ layer HD DVD and Blu-ray discs are also not very interesting. More than likely, these efforts are merely for marketing purposes: to show that HD DVD can match Blu-ray, and that Blu-ray has a bright future. Unfortunately, these are both specious arguments, and it is best to judge them on their initial implementations.
One of the few alternative approaches that looks very promising uses co-linear holography on an optical disc. The advantage is that it can record multiple bits in the same area (volume actually) at the same time, so it scales much better with both density and speed. It may be a ways off yet, but one thing is for certain: an optical disc can only spin so fast, and recording bits one at a time simply doesn't scale well.
Blu-ray is the best we can hope for it the near future. From a data storage perspective, it is far superior to HD DVD, and will remain so until they are both obsolete.
There is no need for levitation when the system is idle, and in that case it will rest on wheels, just like the train. Wear and tear is not an issue, as the wheels must only accommodate very low speed operation, and are rarely in use.
There may be issues with the idea, but Earnshaw's theorem is completely irrelevant. The mere mention of it suggests complete ignorance of the technology involved. That is not unexpected, since the article makes no mention of Inductrack, but why continue with this?
Slashdot Mirror
I don't "overlook" that at all; since that's what real-world hardware is like, real-world fault tolerant software needs to deal with it.
Not at all. No one outside of ignorant home users tolerate such hardware. In the real world, hardware that ignores cache flushes (ie. does not guarantee the ordering of writes) is absolutely not tolerated. This is not a fault, and it can't be worked around in software.
Data that doesn't ever get written will obviously be lost. However, it appears that ZFS problems with such hardware go beyond that.
Again, you are missing the point, this has nothing to do with unwritten data. Bad hardware as described above can still damage traditional filesystems beyond repair, though usually the corruption is more localized. It's not that it doesn't happen, just that you usually won't find out immediately unless it's catastrophic. By the time you discover it, the errors may have already propagated to your backups as well, and then you are screwed.
I should have known better than to post in an Apple focused thread. It is no use trying to inform the willfully ignorant.
What good is a fault tolerant file system if it isn't tolerant of faults?
This is not about faults; ZFS handles those fine. This is about hardware which behaves badly.
You should not talk so authoritatively when you are so obviously ignorant of the subject. What you are implying, is that a filesystem does not care about the ordering of writes, and that is absolutely absurd. The ordering of writes is more critical for copy on write filesystems like ZFS, but in neither case is your data safe on bad hardware.
Like you point out, there is lot of bad hardware out there. What you overlook is that existing filesystems have no facilities to catch, much less correct such errors or corruption; that is why it is called silent data corruption. Even filesystems like HFS+ only journal metadata, so this is a lot more common than you realize.
ZFS can be improved on misbehaving hardware, but that still won't fix bad hardware. Those improvements will only allow the possibility to recover to a consistent state, but data will still be lost.
This is almost entirely nonsense. I have been following the zfs-discuss list for years, and almost no one has lost data. There have been a few bugs which could in rare cases render your data inaccessible, but they almost always have workarounds, and do get fixed.
The data loss and corruption that the parent is talking about is the fault of crap hardware. In almost every case, USB is involved, or more rarely the lack of ECC ram. It is true that ZFS is less tolerant of bad hardware. Note, faulty good hardware is not considered bad; that is reserved for garbage which (for instance) lies to the OS about flushing the disk cache. With such hardware, it is impossible for any filesystem to function reliably.
USB and Firewire bridges are notorious for this. If you care about your data, you should run the other way if you happen upon one. ZFS works great on good hardware though. With directly attached disks and ECC ram, there is no cause for concern.
AMD has supported nested page tables since the Shanghai series processors.
We should be building IFR style plants even if Polywell fusion works out in the next couple of years.
For no other reason than to have somewhere to dispose of the enormous amount of nuclear waste that has been and continues to be generated from conventional nuclear plants. Almost the entirety of that "waste" can be used as fuel in these reactors, rather than sitting in ponds where it will inevitably bite us in the ass.
The true waste from an IFR is very minimal and easily managed, the plants are passively safe, and the reprocessing is proliferation proof.
I am all for investment in fusion technology, but that money should be better spent. Even if the Tokamak approach can be made to work, it will never be economically viable. It is extremely expensive, and there is no hope of scaling it down; the physics requires such an enormous and complex machine.
In the mean time, there are a number of other very promising approaches which continue to be neglected, and these could be funded at a small fraction of the cost. What is very frustrating is that most of these have been around for a long time, and some were even cancelled so that our futile pursuit of Tokamaks could continue.
It is not that Communism didn't or doesn't work, but rather it has never existed as Marx envisioned. The reform has never succeeded. Communism, like Democracy, exists only as a fantasy.
As an ideal, it is not half bad. Sadly, the worst kind of humans always manage to find a way to ruin things for everyone, regardless of the government.
This is excellent news. As you allude, PRAM will finally make good on the promise of solid state storage. It will allow for both higher reliability and deterministic performance, without the ludicrous internal complexity of Flash based devices.
I can't help but cringe every time I hear the terms Flash and SSD used interchangeably. If anything, the limitations inherent to Flash devices described by the GP mean they have more in common with a hard disk, as they also have an inherent physical "geometry" which must be considered.
PRAM will basically look like a simple linear byte array, without all the nonsense associated with Flash. Even if Flash retains a (temporary) advantage in density, it will never compete with hard disks on value for bulk storage, nor will it ever compete with a proper SSD on a performance basis. It makes for a half-assed "SSD", and I can't wait for it to disappear.
The advantage of a breeder reactor like the IFR is that it uses a fuel cycle in which this is effectively impossible. You would need to run a completely different fuel cycle, and likewise, the type of reprocessing facilities required are completely different. This is a good thing.
If we really need more Plutonium, we should be looking to dismantle our weapons stockpile instead. It is way way beyond what could ever be considered reasonable.
As you suggest, there are now a number meltdown-proof reactor designs. These are not merely engineered with "infallible" safety mechanisms, but are fundamentally meltdown-proof by their very design. As long as the laws of physics hold, which is a reasonably safe assumption, there is no risk of meltdown.
While the Pebble bed reactor is safe though, the nature of the pebbles make for very difficult reprocessing, and otherwise still pose a long term waste management problem.
Nuclear is the clear winner for clean, environmentally friendly energy production, but I would recommend pointing people to the Integral Fast Reactor instead. An added benefit would be that such a design could also solve our current nuclear waste problems, by recycling it for use in such reactors. The true waste after recycling is both very minimal and very short lived by comparison.
Sun has been making quite a bit of noise in the storage architecture world with their use of SSDs as intermediate cache to improve reading and writing speeds.
You are conflating Sun's claims here, as the performance gains from using SSDs in their configuration are not generally applicable to other Flash based systems.
ZFS will use SSDs in two very different ways: as cache(L2ARC) devices, and log devices. The cache devices are for improving read IOPs on a mostly static working data set, and a large Flash-based SSD is fine in this scenario. The log devices are for reducing the latency of synchronous writes, and a small DRAM-based SSD is used in this case.
Apple does say that the 17" MBP has a 60% greater gamut, so perhaps it has been improved. I haven't have a chance to compare though.
The LED displays do have other nice features, and are easy on the eyes. The glossy could be easily fixed with an anti-reflective coating, and as you put it, this should not be a problem on a $3000 machine.
Agreed, they both suck. I think what you are referring to is the anti-reflective film that was often used on CRTs. This provides the advantages of both glossy and matte, with none of the disadvantages. It is strange how people seem to have forgotten all about it.
As a reminder, examine the anti-reflective coating on a good pair of glasses; it is extremely effective. If Apple would simply apply such a coating to their displays, this whole matter could be put to rest. Matte/glossy does not need to be a choice...
Just when wide gamut LCDs are approaching the range of colors once possible on CRTs, Apple has taken yet another step backwards with their new LED backlight displays.
My LED MacBook has a spectacularly bad display, so I went to visit the local Apple store to see if this was typical. Sadly it is, and what's more, it looks like all of Apple's LED displays are vastly inferior to that of my old iMac G5. (which has an S-IPS panel and conventional fluorescent backlight)
Color wise, the LED MacBook Pro and Cinema Display are better than the MacBook, but they are all shamefully bad, and definitely worthy of a "worst in the industry" rating. (at least color-wise)
In this case, powers of 2 are more convenient for both the machine and user, as that is how virtually all software and operating systems display disk usage.
It is bad form to present data in an inconvenient format for the user, and using the SI units does exactly that.
so, this is just as worthless as Gigabyte's i-RAM.
While ZFS was not designed for flash, it is basically a perfect fit, as it is a copy-on-write filesystem.
Richard Bennett could not more obviously present himself as a shill for comcast and the like. The article is complete nonsense, attempting to portray bittorrent as the enemy of the Internet.
This idiot drones on about the "ungentlemanly" conduct of using UDP for such purposes, but conveniently avoids the fact that comcast/sandvine caused this mess by injecting face TCP resets, to break bittorrent's TCP connections. Well, what does he expect would happen?
Obviously, UDP is not a good choice for bulk transfers as it lacks congestion control, but lets be fair about where the fault lies. This is not something that can be worked around at the application level, and after being pushed into this corner, there is little else that can be done to work around their abuses of the TCP protocol.
I can not recommend these books enough. Feynman does a brilliant job of bringing the concepts of physics to life.
All together, they are quite extensive, but the individual topics are brief enough to digest in one sitting. Wether you only have a passing interest in physics, or a graduate degree in the field, you will find that there is much to appreciate in these lectures.
Even for those simply taking physics as requirement, I think that these would give you a real appreciation of the field, and probably make the classes a lot easier at that.
Well, perhaps it is in part my physics background, but I didn't get that impression at all. It is a brilliant idea, and even if you aren't familiar with Dr. Bussard, the man knows what he is talking about. He was simply old, somewhat bitter, and impatient--seemingly with good cause. Sadly, he won't see the results of his endeavors, but the research is solid, and thankfully, the navy is following it up.
In any case it has nothing to do with conspiracy theories or blaming fellow scientists. The fact is, basically nothing aside from Tokamak research is funded at a significant level.
Font rendering on the Mac is great, and one of the primary reasons I stick with the platform. The article you cite is nonsense; if you examine the images, it is clear that sub-pixel rendering isn't even enabled for the Mac. (The pure greyscale font images make that plainly obvious...)
It may not be as readable with 4 point fonts, but that is a better compromise than poorly rendering all fonts.
Perhaps most importantly though, the Mac actually comes with good fonts. I don't know about the other languages, but the Japanese fonts in XP are horrendously bad.
No need to compare with 15k rpm drives; flash disks lose spectacularly to low rpm laptop drives for random write performance. For obvious reasons though, no one ever tests random write performance. Manufacturers also rarely report random write IOPS.
Flash is great, if your disk is basically read-only.
Regular fluorescents also contain mercury (a lot more of it) and they also end up in land fills. While CFLs are relatively newer, most people don't even know that conventional fluorescents contain mercury, so there is obviously a problem. Of course, the problem extends to many other hazardous wastes, such as electronics, NiCd batteries, paint, etc.
Whatever the case, there need to be more incentives to safely dispose of hazardous waste. Disposal should be easily accessible, and the costs should be included in the purchase price, with no exceptions. Someone in government needs to get to work on this, because the problem won't fix itself. As is the case with such externalities, corporations are happy to overlook them to our detriment.
Unless they have found a way to record 100 layers at once, it will take nearly forever to record a disc with this new format. For the same reason, the proposed 3+ layer HD DVD and Blu-ray discs are also not very interesting. More than likely, these efforts are merely for marketing purposes: to show that HD DVD can match Blu-ray, and that Blu-ray has a bright future. Unfortunately, these are both specious arguments, and it is best to judge them on their initial implementations.
One of the few alternative approaches that looks very promising uses co-linear holography on an optical disc. The advantage is that it can record multiple bits in the same area (volume actually) at the same time, so it scales much better with both density and speed. It may be a ways off yet, but one thing is for certain: an optical disc can only spin so fast, and recording bits one at a time simply doesn't scale well.
Blu-ray is the best we can hope for it the near future. From a data storage perspective, it is far superior to HD DVD, and will remain so until they are both obsolete.
There is no need for levitation when the system is idle, and in that case it will rest on wheels, just like the train. Wear and tear is not an issue, as the wheels must only accommodate very low speed operation, and are rarely in use.
There may be issues with the idea, but Earnshaw's theorem is completely irrelevant. The mere mention of it suggests complete ignorance of the technology involved. That is not unexpected, since the article makes no mention of Inductrack, but why continue with this?