Mailing List Archive Mailing List Archive
Richard Pyle sez: > Now, add two more identical computers, any one of which can fully operate > the system. The computer complexity has just increased by a factor of 3, > so the probability of a component failure has tripled. However, since 3 > simultaneous computer failures are required to bring the whole system > down, the probability of a system failure is the cube root of the > probability of a single computer failure (i.e., MUCH less -- assuming the > three computers are truly independant entities). Pretty close, but you are ignoring the (additional) complexity of any hw/sw used to link the three computers. (Yes, this is a nit, the end result is still that the reliability is improved!) For more in-depth treatment of this sort of subject, see the reference: _The Theory and Practice of Reliable Systems_, by Siework and Schwarz (sp?). Side note: I do a lot of reliability analysis type work, and recently I had a chance to apply that to everyday life: the NYC Subway has recently taken to advertising how much more wonderful the subway is these days; one of their ads touted that subway cars now travel approx 50,000 miles without a failure -- which they clearly thought was an amazingly impressive number. Well, applying a little systems analysis, looking at the sheer number of miles and cars traveling on a small portion of the system (the 15 mile line I take everyday), I estimated that even with a 50,000 mile MTBF, the system would see a major failure on just my line alone once a week. This doesn't take into account signals, etc -- and guess what? Actual failure rate that I experience is approximately once a week rush hour failure -- pretty good correlation! (I wrote a letter to the MTA -- the subway people -- and they never responded, but the 50,000 mile ads disappeared within a week...) -frank -- fhd@in*.ne* | There is no excellent beauty that hath not some strangeness in 1 212 559 5534 | the proportion! -- Francis Bacon 1 917 992 2248 | Beauty is the proper conformity of the parts to one another 1 718 746 7061 | and to the whole. -- W. Heisenberg
Navigate by Author:
[Previous]
[Next]
[Author Search Index]
Navigate by Subject:
[Previous]
[Next]
[Subject Search Index]
[Send Reply] [Send Message with New Topic]
[Search Selection] [Mailing List Home] [Home]