Uhhh.. you think tape backups are very cheap these days? Where are you looking.
I see 20/40gb drives for like $500. There is some new iomega deal that is 35g for $299.
Anyway I wouldn't call that "very cheap". Not as bad as 5 years ago but still ridiculous considering. I can buy 10 40g hard drives for the cost of a damn tape drive. Quite annoying.
I really hope someone replies to this an proves me wrong. I'd love to find a reasonably priced 100-200gb backup solution.
Wow. You actually took the time and energy to give me shit about a typo. hmmm
You make my point exactly, respect should be earned. If you read the post I was replying to, you would have seen I was taking issue with blindly giving respect based on age.
I totally agree with you. Respect others. Point well taken. My comment apply more to expecting to be regarded more highly just because of your age. On the giving respect you are correct. No down side.
I agree there are issues in the areas you mentioned. However I take issue with the concept of respecting elders. In my opinion (amoung adults) age means very little. Experience, yes! Capability, yes! Wisdome, Yes! These are all good reasons to show respect. But age it's self means nothing.
Respect is something that has to be earned and I know 25 year olds that are twice as capable, wise and worthy of respect than some 50 year olds.
The concept made some sense in ancient tribal society where it was a pretty good bet that the old people killed more buffalo so probably were better at it.
How about I respect the 30 year old, hard working, responsible father of two instead of the 40 year old dead-beat-dad guy who doesn't pay his bills and often mises work because he's two hung over.
In general this doesn't exactly apply to children but still it is more a "respect adults" thing than respect the older people more thing.
I love the "Barring religious reasons". Exactly why should your specific religion give you extra rights over any one else. If you allow something for someone's religion fine, but in my book you really need to allow that same thing for everyone.
Oh, I almost forgot. (ways you can impact me are so easy to think of)
What if I lived in Mississippi and those likely to by my house are white-supremeists (lot's o' them in mississippi) You'll need to leave if your black because that effects my property value. And if your not you better not bring any of your mexican friends over.
Oh... and stop driving your car. People had to move to the city because of the smog.
I own a house. If you are so close to being forclosed on that some aluminum paneling will collapse the value then you have other problems.
If you plan to LIVE in your house you will be more negativly effected by an increase in value as it increases your taxes. If you plan to just buy the house, tell everyone else what to do to increase value and then sell to turn a profit, you really arn't part of the neighborhood then are you? If you dislike my pink house, come chat with me and I'm sure we can come up with a color that works.
Fine then DO not smoke, it raises my health insurance.
Do not own a sports car, it raises my auto insurance
Do not have more than 1 child because i'll have to pay for it.
Don't skydive, you die and effect my life insurance rates
Devaluing a homes value doesn't directly implact those who LIVE nearby only those who sell and leave. Those who stay would actually benefit by a decrease in taxes. Oh... and you can't own a dog because he just might bark and wake me up hence directly *effecting me*. Really, Just mind your own buisness and increase the value of your home by making it nice and participating in your community/local school.
And your nice new addition with flower beds hurt me because it raises my property value and therefore my taxes. Taxes are of MUCH more interest to folks who LIVE in a neighborhood and arn't just trying to turn a buck. You are leaving, why should anyone care how much you get for your house. If you are staying and think blue looks better than pink, come talk to me, we can work something out.
I can come up with a million ways you effect me. You might raise my health insurance rates by smoking. You might have 10 kids that I have to help pay to educate. Point is, it's none of my buisness if you smoke or have 10 kids even though it might impact me.
Fine. So be it. His paitjob is noneof my buisness.
Really if you want to think that way. I hold you responsible for planting flower because it raises my property value and therefore my taxes. Taxes are of MUCH more concern to citizens who LIVE in the neighborhood rather than someone trying to turn a buck. Why should I care how much you can sell your house for? You are leaving. If you are staying and think blue is nicer to look at than pink, come chat with me and I'm sure we can work something out.
I'm well aware that the above opinion is a bit over the top. My point is if you are going to make a fuss over decreasing property value you can make the same fuss about raising the taxes or complaining that because I skydive and some skydivers get hurt it raises your insurance rates.
HOAs are viruses. Some homeowners 30-40 years ago created them and now you can't buy a house without agreeing to them. Hell I didn't even get a full copy of mine but HAD to sign it to buy my house. There is nobody in charge of them and no record of which houses are under the same HOA so you can't get together with your neighbors to change them. My in-laws have a provision in theirs that indicate that they have to have the interrior lights of their house off by 10:00pm.
If a subject is SOO important that it is of public interest then make a LAW about it. At least laws must be overseen by a government body with a known procedure to dispute/change them. If you like HOAs then go life in North Korea. They are right up the communists alley.
Uhhh. Frankly I couldn't care less that it uses special solution that's more expensive. I have ~1400 square feet of hardwood floors and two children under 4 who manage to make it require almost daily mopping. There are a lot of other things I'd like to do than spend the time it takes to vacuum and mop that much space. This sounds like a lovely invention that will be cheaper than the cleaning service I've been considering.
However If I only had one room I probabbly wouldn't do it.
Re:I'm not a Californian
on
Tinfoil Hat House
·
· Score: 4, Insightful
I agree. If it's their property let them be protected from the aliens. Who cares. And don't give me crap about it looking nice for the neighbors. I believe in freedom. If my neighbors want to paint their house with pink polka-dots then so be it, none of my buisness. People spend far too much energy worrying about what other people are doing. The officials should worry about stuff that is actually harming others.
Perhaps you misunderstood me. I wasn't trying to say the author was wrong. The test will fail. I've this behavior in the past. My contention is only that it pretty much doesn't matter to an ordinary application. (proven by the fact that all our computers and servers work just fine) It is mostly an issue with transactional systems and there are solutions to the problem for them.
Yep, but that isn't what happens. On many many systems unless you disable the drives write cache it will not write to disk. You can see this by doing the following. Write two programs for linux, one that sits on computer a and sends data to computer b and expects a respons when it is synced. Program two resides on computer B and recieves the data, writes it and fsyncs then responds.
Get you programs running and pull the plug out of computer b. When it comes back up you will sometimes find that it told computer A it synce the last piece of data but it acutally hadn't.
I work on a transactional messaging system and observed this exact behavior with drives from several vendors. You can disable the physical write cache on the drive and rerun the test and it will work fine. No lost data.
This is probably what the author means by lie. Anyway you have a fine point about the fsync man-page.
The author is specifically talking about the fsync function not the ATA sync command. fsync is an OS call notifying the system to flush it's write caches to the physical device. This writes to the disks write cache but I don't believe it actually issues the sync command to the drive.
In the case of a journaling file system they issue the sync command to the drive to flush the data out.
I work on a block-level transactional system that requires blocks to be synced to the platters. There where two options, modify the kernel to issue syncs to the ata drives on all writes (to the the disk in question) or to just disable the physical write cache on the drive. Turned out to be a touch faster to just diable the cache but the two are effectivly equal.
However drives operate fine under normal conditions, applications write to file systems which take care of forcing the disks to sync. fsync (which the author is talking about) is an OS command and not directly related to the disk sync command.
You are correct. However the article is actually talking about the fact that the data is not on the physical disk when the author fsyncs. That's what I was addressing. And the cache that I'm talking about is the physical cache memory on the drive, not the OS write cache. The "fix" command he mentions that using hdparm is to instruct the drive to disable it's onboard write cache.
What you said is exactly why it doesn't matter to most people and applications. The author was using fsync with the expectation that it would not return until all data was on the platters.
You mention that fsync i fluhing the data to disk. This is close but not entirely true, in linux for example, it is the flushing of the data to the "drive", the OS is satisfied once the data has arrive at the drive, but this is often in the drives physical cache (not the os cache).
There are however certain types of systems that do indeed need to ensure that data is physically written to disk every time. Transactional systems for example. In these cases it is required that either the (Pysical on-board) cache of the drive be disabled or to utilize a disk sub-system with battery-backed cache.
Anyway I think we mostly agree. The author is pretty much having a cow about a known issue that only matters in specialized system and for those there are solutions.
It's not a lie. fsync syncs to a device. The device is a hard drive with a cache.
You'd expect a fsync to complete only when the data is physically written to disk. However usually this is not the case it completes only when it is fully written to the cache on the physical disk.
The downside of this is that it's possible to loose data if you pull the power plug (usually not just by hitting the power switch). However if the disks were to actually commit fully to the physical media on every fsync you would see a very very dramatic performance degredation. Not just a little slower so you look bad in a magazine article but incredibly slow, especially if you are running a database or similar application that fsyncs often.
Server class machines solve this problem by providing battery backed cache on their controllers. This allow the full speed operation by fsyncing only to cache but if power is lost the data is then safe because of the battery.
This doesn't matter too much for the average joe for a number of reasons. First the when the power switch is hit, the disks tend to finnish writing their caches before spinning down. IN the case of a power failure journaled file systems will usually keep you safe (but not always).
This is a big issue however if you are trying to implement an enterprise class database server on everyday hardware.
So turn off the write cache if you don't want it on but don't complain when your system starts to crawl.
So then don't build them you insensitive clod.
Why do people seem to then that the kernel is ONLY for them and their market. Just because there is a driver doesn't mean it needs to bloat your kernel. With simple config options you can build a very small tight kernel.
If anything the extra junk benefits them because the folks developing those drivers are likely to find bugs in the kernel proper.
As I am not writing an article, but a short post, I'll pass on your request for sources. I base my comment (not backed up by the NYTimes research dept) on recent news article from a variety of sources. Should said news prove to be false, I'm sure I'll come back here and print a retraction. (NOT) Feel free to google for it.
An before you call them "my opponent", know that I would be just as annoyed if "LinuxWorld" or IBM funded a study that showed linux ruled. If you "CLAIM" top be an objective analyst, then that is the standard I will hold you to. Since I claim to be "a guy posting on a website"(TM) Please hold me to that standard and I make no claims.
I could be wrong but I don't ever recall slashdot claiming to be an objective news source. Its "news for nerds. stuff that matters" It's biased almost by definition.
It would be like expecting www.goddot.com "News for christians, stuff that prays" to be objective.
Uhhh. Ok. There is a large difference between posting a comment on an entertainment message board and publishing a professional opionion.
Professional integrity applies to ones profession. My profession is not to post on slashdot, her profession is to publish objective opinions on technology. First she participated in research funded by one of the vendors and second she publicly whined about her critics again via her professional identity.
Slashdot Mirror
Uhhh.. you think tape backups are very cheap these days? Where are you looking. I see 20/40gb drives for like $500. There is some new iomega deal that is 35g for $299. Anyway I wouldn't call that "very cheap". Not as bad as 5 years ago but still ridiculous considering. I can buy 10 40g hard drives for the cost of a damn tape drive. Quite annoying. I really hope someone replies to this an proves me wrong. I'd love to find a reasonably priced 100-200gb backup solution.
Wow. You actually took the time and energy to give me shit about a typo. hmmm
You make my point exactly, respect should be earned. If you read the post I was replying to, you would have seen I was taking issue with blindly giving respect based on age.
I totally agree with you. Respect others. Point well taken. My comment apply more to expecting to be regarded more highly just because of your age. On the giving respect you are correct. No down side.
I agree there are issues in the areas you mentioned. However I take issue with the concept of respecting elders. In my opinion (amoung adults) age means very little. Experience, yes! Capability, yes! Wisdome, Yes! These are all good reasons to show respect. But age it's self means nothing.
Respect is something that has to be earned and I know 25 year olds that are twice as capable, wise and worthy of respect than some 50 year olds.
The concept made some sense in ancient tribal society where it was a pretty good bet that the old people killed more buffalo so probably were better at it.
How about I respect the 30 year old, hard working, responsible father of two instead of the 40 year old dead-beat-dad guy who doesn't pay his bills and often mises work because he's two hung over.
In general this doesn't exactly apply to children but still it is more a "respect adults" thing than respect the older people more thing.
Websters defines maturity as "The state or quality of being fully grown or developed."
20-30 years ago people all dressed exactly the same (perhaps a different hue of blue tie), I don't think being a sheep is a sign of maturity.
I love the "Barring religious reasons". Exactly why should your specific religion give you extra rights over any one else. If you allow something for someone's religion fine, but in my book you really need to allow that same thing for everyone.
Oh, I almost forgot. (ways you can impact me are so easy to think of)
What if I lived in Mississippi and those likely to by my house are white-supremeists (lot's o' them in mississippi) You'll need to leave if your black because that effects my property value. And if your not you better not bring any of your mexican friends over.
Oh... and stop driving your car. People had to move to the city because of the smog.
I own a house. If you are so close to being forclosed on that some aluminum paneling will collapse the value then you have other problems.
If you plan to LIVE in your house you will be more negativly effected by an increase in value as it increases your taxes. If you plan to just buy the house, tell everyone else what to do to increase value and then sell to turn a profit, you really arn't part of the neighborhood then are you? If you dislike my pink house, come chat with me and I'm sure we can come up with a color that works.
Fine then DO not smoke, it raises my health insurance.
Do not own a sports car, it raises my auto insurance
Do not have more than 1 child because i'll have to pay for it.
Don't skydive, you die and effect my life insurance rates
Devaluing a homes value doesn't directly implact those who LIVE nearby only those who sell and leave. Those who stay would actually benefit by a decrease in taxes. Oh... and you can't own a dog because he just might bark and wake me up hence directly *effecting me*. Really, Just mind your own buisness and increase the value of your home by making it nice and participating in your community/local school.
And your nice new addition with flower beds hurt me because it raises my property value and therefore my taxes. Taxes are of MUCH more interest to folks who LIVE in a neighborhood and arn't just trying to turn a buck. You are leaving, why should anyone care how much you get for your house. If you are staying and think blue looks better than pink, come talk to me, we can work something out.
I can come up with a million ways you effect me. You might raise my health insurance rates by smoking. You might have 10 kids that I have to help pay to educate. Point is, it's none of my buisness if you smoke or have 10 kids even though it might impact me.
Fine. So be it. His paitjob is noneof my buisness.
Really if you want to think that way. I hold you responsible for planting flower because it raises my property value and therefore my taxes. Taxes are of MUCH more concern to citizens who LIVE in the neighborhood rather than someone trying to turn a buck. Why should I care how much you can sell your house for? You are leaving. If you are staying and think blue is nicer to look at than pink, come chat with me and I'm sure we can work something out.
I'm well aware that the above opinion is a bit over the top. My point is if you are going to make a fuss over decreasing property value you can make the same fuss about raising the taxes or complaining that because I skydive and some skydivers get hurt it raises your insurance rates.
Exactly!
HOAs are viruses. Some homeowners 30-40 years ago created them and now you can't buy a house without agreeing to them. Hell I didn't even get a full copy of mine but HAD to sign it to buy my house. There is nobody in charge of them and no record of which houses are under the same HOA so you can't get together with your neighbors to change them. My in-laws have a provision in theirs that indicate that they have to have the interrior lights of their house off by 10:00pm.
If a subject is SOO important that it is of public interest then make a LAW about it. At least laws must be overseen by a government body with a known procedure to dispute/change them. If you like HOAs then go life in North Korea. They are right up the communists alley.
Uhhh. Frankly I couldn't care less that it uses special solution that's more expensive. I have ~1400 square feet of hardwood floors and two children under 4 who manage to make it require almost daily mopping. There are a lot of other things I'd like to do than spend the time it takes to vacuum and mop that much space. This sounds like a lovely invention that will be cheaper than the cleaning service I've been considering.
However If I only had one room I probabbly wouldn't do it.
I agree. If it's their property let them be protected from the aliens. Who cares. And don't give me crap about it looking nice for the neighbors. I believe in freedom. If my neighbors want to paint their house with pink polka-dots then so be it, none of my buisness. People spend far too much energy worrying about what other people are doing. The officials should worry about stuff that is actually harming others.
Perhaps you misunderstood me. I wasn't trying to say the author was wrong. The test will fail. I've this behavior in the past. My contention is only that it pretty much doesn't matter to an ordinary application. (proven by the fact that all our computers and servers work just fine) It is mostly an issue with transactional systems and there are solutions to the problem for them.
Yep, but that isn't what happens. On many many systems unless you disable the drives write cache it will not write to disk. You can see this by doing the following. Write two programs for linux, one that sits on computer a and sends data to computer b and expects a respons when it is synced. Program two resides on computer B and recieves the data, writes it and fsyncs then responds.
Get you programs running and pull the plug out of computer b. When it comes back up you will sometimes find that it told computer A it synce the last piece of data but it acutally hadn't.
I work on a transactional messaging system and observed this exact behavior with drives from several vendors. You can disable the physical write cache on the drive and rerun the test and it will work fine. No lost data.
This is probably what the author means by lie. Anyway you have a fine point about the fsync man-page.
The author is specifically talking about the fsync function not the ATA sync command. fsync is an OS call notifying the system to flush it's write caches to the physical device. This writes to the disks write cache but I don't believe it actually issues the sync command to the drive.
In the case of a journaling file system they issue the sync command to the drive to flush the data out.
I work on a block-level transactional system that requires blocks to be synced to the platters. There where two options, modify the kernel to issue syncs to the ata drives on all writes (to the the disk in question) or to just disable the physical write cache on the drive. Turned out to be a touch faster to just diable the cache but the two are effectivly equal.
However drives operate fine under normal conditions, applications write to file systems which take care of forcing the disks to sync. fsync (which the author is talking about) is an OS command and not directly related to the disk sync command.
You are correct. However the article is actually talking about the fact that the data is not on the physical disk when the author fsyncs. That's what I was addressing. And the cache that I'm talking about is the physical cache memory on the drive, not the OS write cache. The "fix" command he mentions that using hdparm is to instruct the drive to disable it's onboard write cache.
What you said is exactly why it doesn't matter to most people and applications. The author was using fsync with the expectation that it would not return until all data was on the platters.
You mention that fsync i fluhing the data to disk. This is close but not entirely true, in linux for example, it is the flushing of the data to the "drive", the OS is satisfied once the data has arrive at the drive, but this is often in the drives physical cache (not the os cache).
There are however certain types of systems that do indeed need to ensure that data is physically written to disk every time. Transactional systems for example. In these cases it is required that either the (Pysical on-board) cache of the drive be disabled or to utilize a disk sub-system with battery-backed cache.
Anyway I think we mostly agree. The author is pretty much having a cow about a known issue that only matters in specialized system and for those there are solutions.
It's not a lie. fsync syncs to a device. The device is a hard drive with a cache.
You'd expect a fsync to complete only when the data is physically written to disk. However usually this is not the case it completes only when it is fully written to the cache on the physical disk.
The downside of this is that it's possible to loose data if you pull the power plug (usually not just by hitting the power switch). However if the disks were to actually commit fully to the physical media on every fsync you would see a very very dramatic performance degredation. Not just a little slower so you look bad in a magazine article but incredibly slow, especially if you are running a database or similar application that fsyncs often.
Server class machines solve this problem by providing battery backed cache on their controllers. This allow the full speed operation by fsyncing only to cache but if power is lost the data is then safe because of the battery.
This doesn't matter too much for the average joe for a number of reasons. First the when the power switch is hit, the disks tend to finnish writing their caches before spinning down. IN the case of a power failure journaled file systems will usually keep you safe (but not always).
This is a big issue however if you are trying to implement an enterprise class database server on everyday hardware.
So turn off the write cache if you don't want it on but don't complain when your system starts to crawl.
So then don't build them you insensitive clod. Why do people seem to then that the kernel is ONLY for them and their market. Just because there is a driver doesn't mean it needs to bloat your kernel. With simple config options you can build a very small tight kernel.
If anything the extra junk benefits them because the folks developing those drivers are likely to find bugs in the kernel proper.
HaHa! I almost fell out of my chair laughing. I had to forward your post to some co-workers
As I am not writing an article, but a short post, I'll pass on your request for sources. I base my comment (not backed up by the NYTimes research dept) on recent news article from a variety of sources. Should said news prove to be false, I'm sure I'll come back here and print a retraction. (NOT) Feel free to google for it.
An before you call them "my opponent", know that I would be just as annoyed if "LinuxWorld" or IBM funded a study that showed linux ruled. If you "CLAIM" top be an objective analyst, then that is the standard I will hold you to. Since I claim to be "a guy posting on a website"(TM) Please hold me to that standard and I make no claims.
I could be wrong but I don't ever recall slashdot claiming to be an objective news source. Its "news for nerds. stuff that matters" It's biased almost by definition.
It would be like expecting www.goddot.com "News for christians, stuff that prays" to be objective.
Not really a good comparison.
Uhhh. Ok. There is a large difference between posting a comment on an entertainment message board and publishing a professional opionion.
Professional integrity applies to ones profession. My profession is not to post on slashdot, her profession is to publish objective opinions on technology. First she participated in research funded by one of the vendors and second she publicly whined about her critics again via her professional identity.