Final settlement - HDD status and SMART technology and failure

Это окончание невероятной информации про состояние жёстких дисков.

...

the total results of the relationship between MTBF for the entire period of operation to this parameter for the first year may look as follows:

The calculations show that to estimate the average time between failures for three years of disk drive operation, you need to multiply the initial indicator for the first year (for the same switch-on time and ambient temperature) by a factor of 1.56. Similarly, one can calculate the average time between failures for five and ten years, multiplying the value of the initial parameter for the first year by 1.76 and 1.95, respectively.

Final settlement

Based on all the coefficients calculated above, we can convert the mean time to failure specified by Seagate (for the first year of operation, with 2,400 working hours per year and a switch on duration of 100%) into the average time between failures for the drive installed in the final device user and working at a specific ambient temperature with a certain duration of inclusion. After that, you can also estimate the average time between failures for the entire lifetime of the drive.

Below is an example of calculating the average time between failures during the first year and the entire lifetime for a drive operating for 2,400 hours per year at a temperature of 34 ° C, a switch-on time of 30% and calculated for a service life of 5 years.

As a final example, consider a Seagate drive with one hard drive, for which the mean time to failure during the first year is 444,000 hours. Suppose it is installed in a household electronic device, it is used 2,920 hours per year (8 hours per day, 7 days per week) at an ambient temperature of 42 ° C and a switch-on duration of 5%.

Conclusion

The method described above allows the use of Seagate laboratory test data to assess the reliability of drives installed in desktop computers and consumer electronic devices that operate in a “real world” environment. In short, this method boils down to the following:

• Using Weibull analysis or RDT / FRDT test statistics, estimate Weibull parameters for lab conditions.
• Using the WeiBayes technique, analyze test data for a specific type of drive; Determine the values ​​of the average failure rate AFR for the first year and the mean time to failure during the RDT tests.
• Correct the results obtained taking into account the difference between the RDT laboratory conditions and the actual temperature and on-time at which the drives installed in the end-user equipment will work.
• Based on the cautious assumption that after the first year of operation, the failure rate will remain constant, roll over the reliability assessment for the first year to a three- and ten-year service life of the drive.

In conclusion, we note that this method makes it possible to mathematically use the results of Seagate laboratory tests to assess the reliability of drives installed in consumer electronic devices.

Notes:
1 SuperSmith, Fulton Findings, Weibull WinSMITH and WinSMITH are registered trademarks of Fulton Findings (1251 W. Sepulveda Blvd., # 800, Torrance, CA 90502, USA).
2 Abernethy, Dr. Robert B., The New Weibull handbook, Second Edition, author's edition, 1996, chapter 5.
3 Abernethy, Dr. Robert B., The New Weibull handbook, Second Edition, author's edition, 1996, Appendix D.
4 To compensate for the uncertainty in estimating parameters on Weibull due to limited drive time, you can increase the confidence level during the RDT tests.
5 Nelson, Wayne, Applied Life Data Analysis, John Wiley & Sons, 1982.

Продолжение:

Часть 1 HDD status and SMART technology and failure prediction. G-sensor in HDD. Types of malfunctions HDD.
Часть 2 Learning Smart to predict hard drive failures - HDD status
Часть 3 Final settlement - HDD status and SMART technology and failure