Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uflorida!gatech!rutgers!att!alberta!idacom!danny
From: danny@idacom.UUCP (Danny Wilson)
Newsgroups: comp.sys.apollo
Subject: Re: MTBFs of Apollos
Summary: MTTF information
Message-ID: <639@idacom.UUCP>
Date: 28 Apr 89 19:48:42 GMT
References: <5105@ncrwic.Wichita.NCR.COM>
Organization: IDACOM Electronics Ltd., Edmonton, Alta.
Lines: 41

In article <5105@ncrwic.Wichita.NCR.COM>, rblinne@ncrwic.Wichita.NCR.COM (Richard Blinne) writes:
>                    Does anyone know the MTBFs for Apollo 3000, 3500,
> 4000, and 4500s.  Thank you.

I would also like to know this information and if possible, a comparision
between the above Apollo nodes and the reliability information for
Sun.

BTW, according to reliability theory, MTBF (mean time Between failures)
is not as meaningful as MTTF (Mean time to failure).


             ---------------          ------------------
             |             |          |
             |<-- MTTF --->|<- MTTR ->|
             |             |          |
     --------|             |----------|

              <------ MTBF ---------->

Therefore the MTBF actually should include the MTTR (mean time 
to repair). Manufacturers that are not aware seldom make this
distintion, and when quoting MTBF information they actually are
telling you MTTF!!

In the exciting world :) of reliability theory (and for those
military type guys), MTTF is the measure of choice. i.e. how
long will it run.

eg. A system could have a MTBF of 50 000 hours, however, it is
a bad system if the MTTR is actually 48 000 hours!!

Caveat Computor



-- 
Danny Wilson
IDACOM Electronics		danny@idacom.uucp
Edmonton, Alberta		alberta!idacom!danny
C A N A D A