mtbf简介（MTBF profile）.docVIP

mtbf简介（MTBF profile） With the widespread use of servers, higher requirements for server reliability have been put forward. The so-called reliability, that is, the product in the specified conditions and within the specified time to complete the function of the specified function; on the contrary, the product or part of it can not or will not be able to complete the specified function is a fault. To sum up, the product failure is less reliable, and the ratio of the total number of failures to the total number of units of life is called rate (Failure), which is often expressed as lambda. For example, 100 hard drives are running, 2 faults in a year, and the failure rate of each hard disk is 0.02 times per year. When the life of the product obeys exponential distribution, the reciprocal of its failure rate is called MTBF Time Failures (Mean Between). That is: MTBF=1/ lambda I recently saw a WD Caviar RE2 7200 RPM drive that can be used on servers, MTBF for up to 1 million 200 thousand hours, and 5 years warranty. 1 million 200 thousand hours for about 137 years, not to say that each kind of hard disk can work for 137 years without trouble. By MTBF=1/ lambda, lambda =1/MTBF=1/137 years, that is, the average annual failure rate of the hard disk is about 0.7%, and in one year, 7 of the 1000 hard disks on the average will be out of order. Pictured above is the famous bathtub curve, with the left diagonal portion having an early failure rate, and the failure rate is generally higher and decreases rapidly over time. The middle part of the curve is the service life, and the failure rate is generally low and basically fixed. The most right part is the depletion period, and the failure rate increases rapidly. Electronic products manufacturers usually test, aging, screening and other means to eliminate the early failure as possible, and then available to customers. When the service life will be completed, the product is about to enter a period of high failure, it needs to be scra