Low pressure effects on Hard Disk Drive Performance/Health
Posted: Sat Mar 14, 2015 10:10 pm
Hello,
So, today was my local science fair and I presented my project "Low pressure effects on Hard Disk Drive Performance and Health". To be completey honest my display board was mediocre but my verbal defense and knowledge of my project was "impress" (how the judges put it). I was greatly disappointed to find no award was given for months of work... I would like to present my project to everyone on this great forum and would like to thank everyone who takes the time to read it. I would greatly appreciate constructive criticism and brutal honesty on my project and improvements that could be made. Thank you.
Low Pressure Effects on Hard Disk Drive Performance and Health
Abstract
Objective: Will an HDD cease to function when exposed to low pressures? The intention is to find the correlation between low pressure environments effects on Hard Disk Drive performances and health
Methods: Initially test the performances and health of 4 HDD's. Save files ( .jpeg, .mp4, .txt) onto HDD. Expose 3 HDDs to low pressure environments for lengths of time. Now test these 4 HDD's. View .jpeg, .mp4, and .txt files. Inspect files. Compare initial and final test results.
Results: No abnormalities in the image, sound, and text files. Generally, HDDs exposed to low pressure had a similar trend in certain values but eventually there values almost went back to normal.
Conclusion: My results didn't support my hypothesis because I believed the HDDs would experience no changes in performance and health but my results were that the low pressure changed attributes but the change was temporary. Some big obstacles for this experiment were obtaining materials for the experiment, the scrutiny required to connect each HDD, and the small sample size. An extension would be to use different HDDs (different capacities, manufacturers), also exposing these to low pressure for a much longer period of time and using a bigger sample size.
Hypothesis
If a Hard Disk Drive is exposed to low pressure for an extended period time then the performance, health, and stability of the Hard Disk Drive will not change.
Background Research: Low Pressure effects on HDD Performance and Health
Have you ever wondered what happens to a Hard Disk Drive when its surroundings are out of the ordinary? To be more specific, what about a Hard Disk Drive in a high altitude environment? The difference between the environments of sea level and high altitudes would be pressure difference (and temperature decrease but we will ignore that in order to isolate the pressure difference and really capture the effects of it). The higher the altitude, the lower the pressure. So we simulate these surroundings and investigate how they factor into a HDD’s performance and health and anticipate HDD failure. This study will either give us freedom for where we take our HDD or restrict where we take them, all while keeping in mind the HDD performance and health.
There are many varying types of Hard Disk Drives that are used in our everyday lives, but in this project I will specifically be looking at 4 used mechanical Seagate Barracuda 7200 SATA, with a 80 GB capacity. This HDD is a very typical type of HDD for domestic and commercial use. One advantage of these HDD’s being used is that it eliminates the fear of a defective or incorrectly manufactured HDD, since these HDD’s have been used for a few years. The most common structure for a HDD contains a head, spindle, platter, actuator arm, actuator axis, and actuator. In the platter are tracks and sectors, which store the information while the head reads/writes on the platter. This head reads/writes on the surface of the disk with only nanometers of an air cushion between them. The flying heads need to be at its appropriate flying height in order to operate correctly, and these systems depend on the correct air pressure to allow that. If the air pressure is lower or higher than its allowed air pressure the results might be catastrophic and permanently damage the HDD such as data loss, head crashes, etc.
A method for inducing low pressure/low air density is using a high performance vacuum pump. The Pro-Craft Vacuum Machine works by having a fixed volume, in this case a bell jar of 9” diameter x 8” high surrounded by grease to increase the suction, and eliminating gas molecules from the sealed volume resulting in a vacuum. Pressure at sea level is 101.325 kPa and the vacuum subtracts 100 kPa resulting in 1.325 kPa. The approximate height for this would be 30, 358 meters. For reference, the troposphere goes up to 8,000 meters, the stratosphere goes up to 50,000 meters, and Mount Everest is 8,848 meters high.
A common method for predicting HDD health and performance is using S.M.A.R.T. Tests (Self-Monitoring, Analysis, and Reporting Technology), and benchmarks. S.M.A.R.T. tests measure a set number of attributes (Airflow temperature Celsius, High Fly Writes, Seek Error Rate, etc ) which with proper interpretation can give an illustration of the HDD’s health. Benchmarks measure read rate which will give us an illustration of the HDD’s performance. These tests give us useful raw values which leave us with the interpretation of the data and determining what the status of the HDD is.
“ One of our key findings has been the lack of a consistent pattern of higher failure rates for higher temperature drives or for those drives at higher utilization levels.” ( Eduardo Pinheiro, Wolf-Dietrich Weber and Luiz Andre Barroso Google Inc.) We can conclude that HDD’s are sometimes unpredictable with their health and performance status. “This result suggests that SMART models are more useful in predicting trends for large aggregate populations than for individual components.” ( Eduardo Pinheiro, Wolf-Dietrich Weber and Luiz Andre Barroso Google Inc.) This proves that SMART models are appropriate tests for predicting trends for groups of HDD’s.
Typical reasons for HDD failure are heat, physical damage, power surges, water damage, corrupt files, and human error. Three umbrella categories are mechanical failure, electrical failure, and logical failure. Mechanical failure examples would include wear and tear, head crashes, etc. Warning signs for these might be strange mechanical noises being produced by the HDD. Consequences of mechanical failures include permanent data loss. Electrical failures are most common and some relevant examples are unsupportable heat surroundings. When electrical failures occurs communication between the computer system and HDD will deteriorate and lack accuracy. Logical failures are the most severe and some examples are gradual deterioration over time (which is inevitable) and physical damage to flying heads. In all three cases, the occurrence is possible and the results are catastrophic and most likely permanent.
To further elaborate on the motivation behind this study, it was conducted to discover the limitations and restrictions (or no restrictions) of personal use HDD’s. This information can be very valuable for the reason that it can guide us towards safer operation practices and prevent permanent data loss and/or irreversible damage resulting in a useless HDD.
Methods
Initially tested 4 clean, working, identical HDD's performances and health with S.M.A.R.T. tests and health diagnostics. Installed into a working computer. When installed, install in relatively electrically static free area. Saved image files (.jpeg) of varying resolutions (1178x1080, 3000x24000, 3072x3072, 4200x5600, 5400x2700, and 5723x5563), music files (.mp4) of varying bit-rates (256kbps and 320 kbps), and a text file (.txt) of size (66.2 kB). Saved these onto an additional storage device. Set up High Performance Vacuum Pump and secure the perimeter of the bell jar with grease to ensure a tight seal. Then exposed 3 HDD to a low pressure environment. HDD A was not exposed to a low pressure environment. HDD B was exposed for 30 minutes in a low pressure vacuum of -100 kilo-pascals. HDD C was exposed 60 minutes in a low pressure vacuum of -100 kilo-pascals. HDD D was exposed for 120 minutes in a low pressure vacuum of -100 kilo-pascals.
Now the 4 HDD's performance and health were tested with S.M.A.R.T. tests and health diagnostics. Saved results. Viewed .jpeg files of varying resolutions (1178x1080, 3000x24000, 3072x3072, 4200x5600, 5400x2700, and 5723x5563), .mp4 files of varying bitrates (256kbps and 320 kbps), and a .txt file of size (66.2kB). Inspected for abnormalities or corruptions. Compared initial S.M.A.R.T. tests and benchmark results with final S.M.A.R.T. tests and benchmark results.
Materials
4 near-identical SATA Hard Disk Drives
a working desktop computer
precision screwdriver
clean, static electricity free area (avoid carpeted areas)
a low pressure vacuum
grease
Results
From our graphs and charts, we see that our data follows trends in some attributes and does not follow trends in other attributes. For Airflow Temperature Range, HDD A’s range follows a downward curve, while HDD B, HDD C, and HDD D follow an upward curve. Hardware ECC Recovered Value’s data does not follow a consistent trend, but after the HDD has been exposed, generally the HDD’s Hardware ECC Recovered Value experiences curvature, then restores to a similar initial value. For Temperature Raw Value (Celsius), HDD A experiences a linear decrease and has its Raw Value deviate the most out of all the HDD tested. HDD B, HDD C, and HDD D experience either positive or negative curvature then restore back to a similar initial value. For Hardware ECC Recovered Raw Value, all HDDs come across curvature, either negative or positive, but once again restore back to a similar initial value. For Seek-Error Rate Raw Value, all HDD experience a positive linear trend.
Graphs and Table
http://imgur.com/a/LR6Pj#4
Discussion
Our Airflow Temperature Range, produces a number of interesting trends and results. First and foremost, HDD A (exposed to low pressure environment for 0 minutes) experienced a dip in the values and restored back to a similar initial value, while HDD B, HDD C, and HDD D (exposed to low pressure environment for 30 min, 60 min, 120 min respectively) experienced an increased values (of varying magnitudes) then restoring back to similar initial values. These trends suggest that the HDDs that were exposed had an increase in values immediately after being exposed. Since we know higher values are a bad sign for the HDD in this particular attribute, we can infer that this increase in range for HDD B, C, and D is a bad indicator for the HDD health/performance. But since this negative deviation is only temporary, there are no serious consequences for the exposure of a low pressure environment, regarding Airflow Temperature Range.
For Hardware ECC Recovered Values (accuracy of data passing through HDD), lower exposure times experience the most dramatic curvature, except for HDD A which does not vary much since it was not exposed. Once again, the values return to similar initial values after experiencing curvature. Since, lower values (concerning values) are bad and higher values are good, we can assume that when exposed to low pressure environment for 30 minutes the HDD goes through an increase in accuracy of data passing through the HDD. These changes in accuracy of data are temporary, and therefore makes the exposure time irrelevant in the long run.
In Temperature Raw Value, lower raw values are a good sign of health so HDDs which experience positive curvature are in trouble immediately after exposure ( like HDD B, HDD C). HDD A has a negative linear slope which indicates a good sign of health when not exposed to a low pressure environment. HDD D has a negative curvature and ends at a lower similar initial value telling us that the effects of low pressure for 120 minutes had a positive effect on the HDD. All HDDs (including those exposed and those not exposed) end at a better sign of health for Temperature Raw Value, saying HDD’s exposed experienced fluctuations in Temperature and the HDD not exposed did not experience fluctuations, but all HDD’s ended with a better sign of health regarding temperature.
For Hardware ECC Raw Values, the fluctuations in the accuracy of passing on data seems arbitrary. Since HDD A seems to follow similar trends to the other HDDs we can conclude that changes in Hardware ECC Raw Values are random and no meaningful conclusion can be determined. So, low pressure exposure has no effect on Hardware ECC Raw Value.
In Seek-Error Rate, all HDD’s follow a linear trend. Since all HDD’s follow the same trend including those not exposed and those exposed, we can determine that low pressure has no effect on Seek-Error Rate.
Keeping all this in mind, certain aspects of my hypothesis were correct and incorrect. I believed that if a HDD is exposed to a low pressure environment for an extended period of time then the performance and health will experience no changes. It is correct for the reason that in the end no permanent damage was done to the HDDs, saying that the performance and health remained relevantly unchanged comparing the initial to the final time of when the data was extracted. It is incorrect for the reason that some of the HDDs attributes experienced fluctuations immediately after being exposed but after settling and not being in use for a day, it gave the HDD time to “normalize” or restore to its similar initial value.
Certain techniques in this study lacked precision and consistency. A strong suggestion for improvement would concern mostly the HDDs lifetime, exposed environment, and similar daily use. For a study similar to this, we would like to have the HDD’s as identical as possible to eliminate as many errors as possible. So, lifetime, exposed environment, and daily use should all be as alike as possible. It would be preferred for these HDD’s to be used for at least 6-12 months to eliminate the factor of defective HDDs. Another strong suggestion would be the sample size. The sample in this study is relatively low, so it would be best to have more than 10 identical HDDs. Another improvement could be a cleaner environment for the HDDs to settle while they are not in use and an enormous amount of attention to carefulness when handling the HDDs.
Taking this study a step further could mean inducing low pressure with different methods, taking HDDs to actual low pressure environments ( high altitudes ), running the HDD while in the low pressure environment, and changing other factors of the environment. With all these variables, still testing HDD health and performance.
Bibliography
Pinheiro, Eduardo, Wolf-Dietrich Weber, and Luis Andre Barroso. "Failure Trends in a Large Disk Drive Population." Google Research (2007). Google. Google. Web. 18 Feb. 2015. <http://static.googleusercontent.com/med ... ilures.pdf>.
"Meet Your Hard Drive." Meet Your Hard Drive. IBM (International Business Machines Corporation). Web. 11 Mar. 2015. <http://www.research.ibm.com/research/gmr/basics.html>.
Allen, Bruce. "Monitoring Hard Disks with SMART." Linux Journal (2004). Linux Journal. Linux. Web. 27 Feb. 2015. <http://www.linuxjournal.com/article/6983>.
"How A Hard Drive Works." How A Hard Drive Works. ACS Data Recovery. Web. 1 Mar. 2015. <http://acsdata.com/how-a-hard-drive-works.htm>.
"Mechanical Hard Drive Failure." Data Failures. Compu Recovery. Compu Recovery. Web. 1 Mar. 2015. <https://www.compurecovery.com/articles/ ... e-failure/>.
So, today was my local science fair and I presented my project "Low pressure effects on Hard Disk Drive Performance and Health". To be completey honest my display board was mediocre but my verbal defense and knowledge of my project was "impress" (how the judges put it). I was greatly disappointed to find no award was given for months of work... I would like to present my project to everyone on this great forum and would like to thank everyone who takes the time to read it. I would greatly appreciate constructive criticism and brutal honesty on my project and improvements that could be made. Thank you.
Low Pressure Effects on Hard Disk Drive Performance and Health
Abstract
Objective: Will an HDD cease to function when exposed to low pressures? The intention is to find the correlation between low pressure environments effects on Hard Disk Drive performances and health
Methods: Initially test the performances and health of 4 HDD's. Save files ( .jpeg, .mp4, .txt) onto HDD. Expose 3 HDDs to low pressure environments for lengths of time. Now test these 4 HDD's. View .jpeg, .mp4, and .txt files. Inspect files. Compare initial and final test results.
Results: No abnormalities in the image, sound, and text files. Generally, HDDs exposed to low pressure had a similar trend in certain values but eventually there values almost went back to normal.
Conclusion: My results didn't support my hypothesis because I believed the HDDs would experience no changes in performance and health but my results were that the low pressure changed attributes but the change was temporary. Some big obstacles for this experiment were obtaining materials for the experiment, the scrutiny required to connect each HDD, and the small sample size. An extension would be to use different HDDs (different capacities, manufacturers), also exposing these to low pressure for a much longer period of time and using a bigger sample size.
Hypothesis
If a Hard Disk Drive is exposed to low pressure for an extended period time then the performance, health, and stability of the Hard Disk Drive will not change.
Background Research: Low Pressure effects on HDD Performance and Health
Have you ever wondered what happens to a Hard Disk Drive when its surroundings are out of the ordinary? To be more specific, what about a Hard Disk Drive in a high altitude environment? The difference between the environments of sea level and high altitudes would be pressure difference (and temperature decrease but we will ignore that in order to isolate the pressure difference and really capture the effects of it). The higher the altitude, the lower the pressure. So we simulate these surroundings and investigate how they factor into a HDD’s performance and health and anticipate HDD failure. This study will either give us freedom for where we take our HDD or restrict where we take them, all while keeping in mind the HDD performance and health.
There are many varying types of Hard Disk Drives that are used in our everyday lives, but in this project I will specifically be looking at 4 used mechanical Seagate Barracuda 7200 SATA, with a 80 GB capacity. This HDD is a very typical type of HDD for domestic and commercial use. One advantage of these HDD’s being used is that it eliminates the fear of a defective or incorrectly manufactured HDD, since these HDD’s have been used for a few years. The most common structure for a HDD contains a head, spindle, platter, actuator arm, actuator axis, and actuator. In the platter are tracks and sectors, which store the information while the head reads/writes on the platter. This head reads/writes on the surface of the disk with only nanometers of an air cushion between them. The flying heads need to be at its appropriate flying height in order to operate correctly, and these systems depend on the correct air pressure to allow that. If the air pressure is lower or higher than its allowed air pressure the results might be catastrophic and permanently damage the HDD such as data loss, head crashes, etc.
A method for inducing low pressure/low air density is using a high performance vacuum pump. The Pro-Craft Vacuum Machine works by having a fixed volume, in this case a bell jar of 9” diameter x 8” high surrounded by grease to increase the suction, and eliminating gas molecules from the sealed volume resulting in a vacuum. Pressure at sea level is 101.325 kPa and the vacuum subtracts 100 kPa resulting in 1.325 kPa. The approximate height for this would be 30, 358 meters. For reference, the troposphere goes up to 8,000 meters, the stratosphere goes up to 50,000 meters, and Mount Everest is 8,848 meters high.
A common method for predicting HDD health and performance is using S.M.A.R.T. Tests (Self-Monitoring, Analysis, and Reporting Technology), and benchmarks. S.M.A.R.T. tests measure a set number of attributes (Airflow temperature Celsius, High Fly Writes, Seek Error Rate, etc ) which with proper interpretation can give an illustration of the HDD’s health. Benchmarks measure read rate which will give us an illustration of the HDD’s performance. These tests give us useful raw values which leave us with the interpretation of the data and determining what the status of the HDD is.
“ One of our key findings has been the lack of a consistent pattern of higher failure rates for higher temperature drives or for those drives at higher utilization levels.” ( Eduardo Pinheiro, Wolf-Dietrich Weber and Luiz Andre Barroso Google Inc.) We can conclude that HDD’s are sometimes unpredictable with their health and performance status. “This result suggests that SMART models are more useful in predicting trends for large aggregate populations than for individual components.” ( Eduardo Pinheiro, Wolf-Dietrich Weber and Luiz Andre Barroso Google Inc.) This proves that SMART models are appropriate tests for predicting trends for groups of HDD’s.
Typical reasons for HDD failure are heat, physical damage, power surges, water damage, corrupt files, and human error. Three umbrella categories are mechanical failure, electrical failure, and logical failure. Mechanical failure examples would include wear and tear, head crashes, etc. Warning signs for these might be strange mechanical noises being produced by the HDD. Consequences of mechanical failures include permanent data loss. Electrical failures are most common and some relevant examples are unsupportable heat surroundings. When electrical failures occurs communication between the computer system and HDD will deteriorate and lack accuracy. Logical failures are the most severe and some examples are gradual deterioration over time (which is inevitable) and physical damage to flying heads. In all three cases, the occurrence is possible and the results are catastrophic and most likely permanent.
To further elaborate on the motivation behind this study, it was conducted to discover the limitations and restrictions (or no restrictions) of personal use HDD’s. This information can be very valuable for the reason that it can guide us towards safer operation practices and prevent permanent data loss and/or irreversible damage resulting in a useless HDD.
Methods
Initially tested 4 clean, working, identical HDD's performances and health with S.M.A.R.T. tests and health diagnostics. Installed into a working computer. When installed, install in relatively electrically static free area. Saved image files (.jpeg) of varying resolutions (1178x1080, 3000x24000, 3072x3072, 4200x5600, 5400x2700, and 5723x5563), music files (.mp4) of varying bit-rates (256kbps and 320 kbps), and a text file (.txt) of size (66.2 kB). Saved these onto an additional storage device. Set up High Performance Vacuum Pump and secure the perimeter of the bell jar with grease to ensure a tight seal. Then exposed 3 HDD to a low pressure environment. HDD A was not exposed to a low pressure environment. HDD B was exposed for 30 minutes in a low pressure vacuum of -100 kilo-pascals. HDD C was exposed 60 minutes in a low pressure vacuum of -100 kilo-pascals. HDD D was exposed for 120 minutes in a low pressure vacuum of -100 kilo-pascals.
Now the 4 HDD's performance and health were tested with S.M.A.R.T. tests and health diagnostics. Saved results. Viewed .jpeg files of varying resolutions (1178x1080, 3000x24000, 3072x3072, 4200x5600, 5400x2700, and 5723x5563), .mp4 files of varying bitrates (256kbps and 320 kbps), and a .txt file of size (66.2kB). Inspected for abnormalities or corruptions. Compared initial S.M.A.R.T. tests and benchmark results with final S.M.A.R.T. tests and benchmark results.
Materials
4 near-identical SATA Hard Disk Drives
a working desktop computer
precision screwdriver
clean, static electricity free area (avoid carpeted areas)
a low pressure vacuum
grease
Results
From our graphs and charts, we see that our data follows trends in some attributes and does not follow trends in other attributes. For Airflow Temperature Range, HDD A’s range follows a downward curve, while HDD B, HDD C, and HDD D follow an upward curve. Hardware ECC Recovered Value’s data does not follow a consistent trend, but after the HDD has been exposed, generally the HDD’s Hardware ECC Recovered Value experiences curvature, then restores to a similar initial value. For Temperature Raw Value (Celsius), HDD A experiences a linear decrease and has its Raw Value deviate the most out of all the HDD tested. HDD B, HDD C, and HDD D experience either positive or negative curvature then restore back to a similar initial value. For Hardware ECC Recovered Raw Value, all HDDs come across curvature, either negative or positive, but once again restore back to a similar initial value. For Seek-Error Rate Raw Value, all HDD experience a positive linear trend.
Graphs and Table
http://imgur.com/a/LR6Pj#4
Discussion
Our Airflow Temperature Range, produces a number of interesting trends and results. First and foremost, HDD A (exposed to low pressure environment for 0 minutes) experienced a dip in the values and restored back to a similar initial value, while HDD B, HDD C, and HDD D (exposed to low pressure environment for 30 min, 60 min, 120 min respectively) experienced an increased values (of varying magnitudes) then restoring back to similar initial values. These trends suggest that the HDDs that were exposed had an increase in values immediately after being exposed. Since we know higher values are a bad sign for the HDD in this particular attribute, we can infer that this increase in range for HDD B, C, and D is a bad indicator for the HDD health/performance. But since this negative deviation is only temporary, there are no serious consequences for the exposure of a low pressure environment, regarding Airflow Temperature Range.
For Hardware ECC Recovered Values (accuracy of data passing through HDD), lower exposure times experience the most dramatic curvature, except for HDD A which does not vary much since it was not exposed. Once again, the values return to similar initial values after experiencing curvature. Since, lower values (concerning values) are bad and higher values are good, we can assume that when exposed to low pressure environment for 30 minutes the HDD goes through an increase in accuracy of data passing through the HDD. These changes in accuracy of data are temporary, and therefore makes the exposure time irrelevant in the long run.
In Temperature Raw Value, lower raw values are a good sign of health so HDDs which experience positive curvature are in trouble immediately after exposure ( like HDD B, HDD C). HDD A has a negative linear slope which indicates a good sign of health when not exposed to a low pressure environment. HDD D has a negative curvature and ends at a lower similar initial value telling us that the effects of low pressure for 120 minutes had a positive effect on the HDD. All HDDs (including those exposed and those not exposed) end at a better sign of health for Temperature Raw Value, saying HDD’s exposed experienced fluctuations in Temperature and the HDD not exposed did not experience fluctuations, but all HDD’s ended with a better sign of health regarding temperature.
For Hardware ECC Raw Values, the fluctuations in the accuracy of passing on data seems arbitrary. Since HDD A seems to follow similar trends to the other HDDs we can conclude that changes in Hardware ECC Raw Values are random and no meaningful conclusion can be determined. So, low pressure exposure has no effect on Hardware ECC Raw Value.
In Seek-Error Rate, all HDD’s follow a linear trend. Since all HDD’s follow the same trend including those not exposed and those exposed, we can determine that low pressure has no effect on Seek-Error Rate.
Keeping all this in mind, certain aspects of my hypothesis were correct and incorrect. I believed that if a HDD is exposed to a low pressure environment for an extended period of time then the performance and health will experience no changes. It is correct for the reason that in the end no permanent damage was done to the HDDs, saying that the performance and health remained relevantly unchanged comparing the initial to the final time of when the data was extracted. It is incorrect for the reason that some of the HDDs attributes experienced fluctuations immediately after being exposed but after settling and not being in use for a day, it gave the HDD time to “normalize” or restore to its similar initial value.
Certain techniques in this study lacked precision and consistency. A strong suggestion for improvement would concern mostly the HDDs lifetime, exposed environment, and similar daily use. For a study similar to this, we would like to have the HDD’s as identical as possible to eliminate as many errors as possible. So, lifetime, exposed environment, and daily use should all be as alike as possible. It would be preferred for these HDD’s to be used for at least 6-12 months to eliminate the factor of defective HDDs. Another strong suggestion would be the sample size. The sample in this study is relatively low, so it would be best to have more than 10 identical HDDs. Another improvement could be a cleaner environment for the HDDs to settle while they are not in use and an enormous amount of attention to carefulness when handling the HDDs.
Taking this study a step further could mean inducing low pressure with different methods, taking HDDs to actual low pressure environments ( high altitudes ), running the HDD while in the low pressure environment, and changing other factors of the environment. With all these variables, still testing HDD health and performance.
Bibliography
Pinheiro, Eduardo, Wolf-Dietrich Weber, and Luis Andre Barroso. "Failure Trends in a Large Disk Drive Population." Google Research (2007). Google. Google. Web. 18 Feb. 2015. <http://static.googleusercontent.com/med ... ilures.pdf>.
"Meet Your Hard Drive." Meet Your Hard Drive. IBM (International Business Machines Corporation). Web. 11 Mar. 2015. <http://www.research.ibm.com/research/gmr/basics.html>.
Allen, Bruce. "Monitoring Hard Disks with SMART." Linux Journal (2004). Linux Journal. Linux. Web. 27 Feb. 2015. <http://www.linuxjournal.com/article/6983>.
"How A Hard Drive Works." How A Hard Drive Works. ACS Data Recovery. Web. 1 Mar. 2015. <http://acsdata.com/how-a-hard-drive-works.htm>.
"Mechanical Hard Drive Failure." Data Failures. Compu Recovery. Compu Recovery. Web. 1 Mar. 2015. <https://www.compurecovery.com/articles/ ... e-failure/>.