Re: [time-nuts] Meaning of MTBF (was: Reliability of atomic clocks)
On 30 Mar 2016 09:00, "Jay Grizzard" wrote: > > > It get's "interesting" when you look at the MTBF times on hard disks. Some > > of the figures quoted in hours related to an MTBF of over 100 years. From > > what I read before, this was based on you replacing the drive at the end of > > its service life (typically 3 years for consumer drives and 5 years for > > enterprise grade disks). > > I note Seagate have dropped the use of MTBF: > > > > http://knowledge.seagate.com/articles/en_US/FAQ/174791en?language=en_US > > The article you link here actually explains what MTBF on drives is > measuring -- and it has nothing to do with when you replace your drives. That article does not. But I have read articles from other manufacturers where the MTBF was defined in terms of drives being replaced at the end of their service life. Seagate have obviously dropped the use of the term MTBF for hard dusks. Dave. ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Meaning of MTBF (was: Reliability of atomic clocks)
> It get's "interesting" when you look at the MTBF times on hard disks. Some > of the figures quoted in hours related to an MTBF of over 100 years. From > what I read before, this was based on you replacing the drive at the end of > its service life (typically 3 years for consumer drives and 5 years for > enterprise grade disks). So no individual drive was ever expected to last > 100 years, but if you kept replacing the drives ever 3~5 years, the average > time of an unexpected failure would be 100 years. I guess its a bit like a > car - the engine might run for 250,000 miles, but if you never change the > oil or the camshaft belt, it is not going to last. > > I note Seagate have dropped the use of MTBF: > > http://knowledge.seagate.com/articles/en_US/FAQ/174791en?language=en_US The article you link here actually explains what MTBF on drives is measuring -- and it has nothing to do with when you replace your drives. MTBF is basically expressed as "1 failure per N power-on hours". So if you have a MTBF of 100,000 hours and you have 100 drives running continuously, you will (on average) have one failure every ~42 days (1000 hours). If you have 100,000 drives, you'll have (on average) one failure every hour. MTBF does not address the expected life of any specific drive in any way. (It also does not address the bathtub curve that drive failures tend to follow -- there's a high 'infant mortality' rate for new drives, then a number of years of service with a low failure rate, followed by an increase in failure rate after some number of years.) FWIW, there have been a few interesting things published on drive failure rates. One of the most interesting is a study[1] Google published in 2007, which drew some rather unexpected conclusions (e.g. drive temperature is not associated with failure rate, except at the higher ends of the temperature range). Backblaze (a cloud backup provider) also publishes regular reports on drive reliability[2], and have been for a few years now. -j 1. http://static.googleusercontent.com/media/research.google.com/en//archive/disk_failures.pdf 2. https://www.backblaze.com/blog/hard-drive-reliability-q4-2015/ ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Meaning of MTBF (was: Reliability of atomic clocks)
On 28 March 2016 at 00:32, Attila Kinali wrote: > > Yes, the MTBF is a very simplicistic measure and there are a couple > of assumptions in its calculation which do not hold generally (or > rather, it's rather seldom that they hold). It get's "interesting" when you look at the MTBF times on hard disks. Some of the figures quoted in hours related to an MTBF of over 100 years. From what I read before, this was based on you replacing the drive at the end of its service life (typically 3 years for consumer drives and 5 years for enterprise grade disks). So no individual drive was ever expected to last 100 years, but if you kept replacing the drives ever 3~5 years, the average time of an unexpected failure would be 100 years. I guess its a bit like a car - the engine might run for 250,000 miles, but if you never change the oil or the camshaft belt, it is not going to last. I note Seagate have dropped the use of MTBF: http://knowledge.seagate.com/articles/en_US/FAQ/174791en?language=en_US changing to an Annualized Failure Rate (AFR). I don't think Seagate will ever get a real measure of this, as in many cases people are just going to throw a hard disk in the bin if it fails, even if under warranty. In many cases the warranty is with an OEM, so even if you buy a new drive sold originally to Dell, you can't return it unless you are Dell. Also with hard drive capacities growing quite fast, if a drive does fail you will probably chose to replace it with one of higher capacity. Dr. David Kirkby Ph.D CEng MIET Kirkby Microwave Ltd Registered office: Stokes Hall Lodge, Burnham Rd, Althorne, Essex, CM3 6DT, UK. Registered in England and Wales, company number 08914892. http://www.kirkbymicrowave.co.uk/ Tel: 07910 441670 / +44 7910 441670 (0900 to 2100 GMT only please) ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Meaning of MTBF (was: Reliability of atomic clocks)
It's been a while since I designed aerospace hardware, but seems I remember we had both a calculated AND a demonstrated MTBF. Back then we called it Mil-Std 781. (I am sure it morphed into more modern tests). We had both a pre-production qual-test and a production acceptance-test, both required to meet MTBF's, that were run for reliability. I took stock in them; as did others. They did have merit in predicting weak engineering designs catching weak designs during 'life' production. It wasn't 'simplistic' at all. Maybe the military and aerospace world is different from the 'commercial' world. -Don == On Mon, 28 Mar 2016 13:04:23 +0200 Florian Teply wrote: > Am Mon, 28 Mar 2016 01:32:03 +0200 > schrieb Attila Kinali : > > > Yes, the MTBF is a very simplicistic measure and there are a couple > > of assumptions in its calculation which do not hold generally (or > > rather, it's rather seldom that they hold). Yet it gives a number to > > something that is otherwise relatively hard to measure and the number, > > even though flawed, makes it possible to compare different devices > > on their reliability. As this is more a rule of thumb comparison, > > you shouldn't read too much into a 10% difference. Yet a 100% > > difference is significant, no matter which of the assumptions do not > > hold. > > > Umm, well, even a 100% difference still might mean nothing if the > derivation of MTBF between different devices is based on different > assumptions. That both these derivations might be seriously flawed does > not help at all. > Yet, even MIL-Spec parts documentation does rarely contain sufficient > detail to assess the validity of the numbers in a certain application. > > At the very least, one would need to know acceleration factors for the > different failure mechanisms, and shape parameters of the > failure-vs-time plot. This kind of data I wouldn't expect to find > outside the manufacturers premises, and even there it's not likely to > be accessible if it exists at all. > > Best regards, > Florian > ___ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. -- dlewis6767 ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Meaning of MTBF (was: Reliability of atomic clocks)
We make some thick film, plastic molded , plug in attenuators for the cable TV industry. We have been asked on several occasions to provide MTBF data. Being a small company with limited resources, we have never been able to provide that data. The parts we make will easily outlast the equipment they are used in, because such equipment is frequently upgraded. We tell our customers that very few, if any, parts have ever been returned for being defective. While true, it is somewhat misleading. The plug in attenuator is an inexpensive part that is carried in a tech's belt bag by the handful. If he encounters a bad one, he simply removes it, tosses it on the ground, and installs a new one. In recent years the Chinese have taken most of the market away from us with lower cost products. A lot of them use FR4 and chip resistors. I have heard that some OEMs are going away from those because of reduced reliability. I have been seeing some Chinese products that are actually thick film. Most are either copies of ours, or they didn't spend any time to optimize the RF performance. Sometimes, I do see parts that out perform our own. The Chinese also make plug in, molded equalizers, but I have yet to see one that works well. I must say to their credit though, that they have gone from cheap imitations to parts that are actually well designed and built. AT least some of them are. On Mon, Mar 28, 2016 at 6:04 AM, Florian Teply wrote: > Am Mon, 28 Mar 2016 01:32:03 +0200 > schrieb Attila Kinali : > > > Yes, the MTBF is a very simplicistic measure and there are a couple > > of assumptions in its calculation which do not hold generally (or > > rather, it's rather seldom that they hold). Yet it gives a number to > > something that is otherwise relatively hard to measure and the number, > > even though flawed, makes it possible to compare different devices > > on their reliability. As this is more a rule of thumb comparison, > > you shouldn't read too much into a 10% difference. Yet a 100% > > difference is significant, no matter which of the assumptions do not > > hold. > > > Umm, well, even a 100% difference still might mean nothing if the > derivation of MTBF between different devices is based on different > assumptions. That both these derivations might be seriously flawed does > not help at all. > Yet, even MIL-Spec parts documentation does rarely contain sufficient > detail to assess the validity of the numbers in a certain application. > > At the very least, one would need to know acceleration factors for the > different failure mechanisms, and shape parameters of the > failure-vs-time plot. This kind of data I wouldn't expect to find > outside the manufacturers premises, and even there it's not likely to > be accessible if it exists at all. > > Best regards, > Florian > ___ > time-nuts mailing list -- time-nuts@febo.com > To unsubscribe, go to > https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts > and follow the instructions there. > ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
Re: [time-nuts] Meaning of MTBF (was: Reliability of atomic clocks)
Am Mon, 28 Mar 2016 01:32:03 +0200 schrieb Attila Kinali : > Yes, the MTBF is a very simplicistic measure and there are a couple > of assumptions in its calculation which do not hold generally (or > rather, it's rather seldom that they hold). Yet it gives a number to > something that is otherwise relatively hard to measure and the number, > even though flawed, makes it possible to compare different devices > on their reliability. As this is more a rule of thumb comparison, > you shouldn't read too much into a 10% difference. Yet a 100% > difference is significant, no matter which of the assumptions do not > hold. > Umm, well, even a 100% difference still might mean nothing if the derivation of MTBF between different devices is based on different assumptions. That both these derivations might be seriously flawed does not help at all. Yet, even MIL-Spec parts documentation does rarely contain sufficient detail to assess the validity of the numbers in a certain application. At the very least, one would need to know acceleration factors for the different failure mechanisms, and shape parameters of the failure-vs-time plot. This kind of data I wouldn't expect to find outside the manufacturers premises, and even there it's not likely to be accessible if it exists at all. Best regards, Florian ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
[time-nuts] Meaning of MTBF (was: Reliability of atomic clocks)
Salut Alan, On Sun, 27 Mar 2016 15:52:00 +0100 "Alan Melia" wrote: > I am out of the business now, well retired, so my opinion carries > little weight, > :-)) but for whatever it does, my thought is that MTBF is a pretty useless > parameter in general. This is a relatively low volume unit manufactured by a > variety of different firms with each their opinion on the best optimum. > > The statistical base to MTBF is faulty and in my opinion its only use is to > indicate where a design might be improved by changing the component mix. The > actual value that falls out of the end of the calculation for a desgn is > completely meaningless, but the non-tech bean-counters wanted a way to > justify more expensive designs, and the purchase of expensive kit. Yes, the MTBF is a very simplicistic measure and there are a couple of assumptions in its calculation which do not hold generally (or rather, it's rather seldom that they hold). Yet it gives a number to something that is otherwise relatively hard to measure and the number, even though flawed, makes it possible to compare different devices on their reliability. As this is more a rule of thumb comparison, you shouldn't read too much into a 10% difference. Yet a 100% difference is significant, no matter which of the assumptions do not hold. I my case here, I use the MTBF as a stand in for a more general reliability probability density function, a term which might confuse more than clarify in the question asked. Attila Kinali Attila Kinali -- Reading can seriously damage your ignorance. -- unknown ___ time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts and follow the instructions there.
