[SATLUG] Average lifespan of an Adaptec SCSI card?

Alan Lesmerises alesmerises at satx.rr.com
Wed Jul 22 23:28:56 CDT 2009


Brad Knowles wrote:
> on 7/22/09 8:56 PM, Alan Lesmerises said:
>> In the case of solid-state electronics, it has long been known that 
>> the bulk of the failures occur in a pattern known as "infant 
>> mortality" (and the name does come from human death studies).  That 
>> means that failures occur with decreasing frequency with time, and 
>> the longer it lasts, the longer it's going to last.  That being said, 
>> it they are subjected to some external factor (excessive overheating, 
>> vibration, etc.), failures can also be externally induced.
> In my experience, it is more typically called a "bathtub curve", 
> because it looks like a bathtub in profile -- high on one end, low in 
> the middle, and then climbing back to high again at the end.
>
> Planned obsolescence is a very real principle that is used by most 
> companies these days when they design their products, such that they 
> really do tend to start failing a little after the warranty has 
> expired.  Yes, the Engineers have gotten good enough with their design 
> tradeoffs and other techniques, that they can come statistically quite 
> close to whatever warranty policies that the business types want to 
> implement.
>
> Of course, there are plenty of other factors to be considered, but 
> that does describe what I've seen as an overall trend.
Yes, that 'climbing back' as you call it _can_ happen, and it is 
referred-to as "wear-out".  However, there needs to be some physics 
behind that wear-out.  For example, the tires on your car wear out 
because the rubber is worn away during its life.  Similarly, mechanical 
parts that experience some sort of physical contact with other parts 
will slowly (or maybe quickly) be worn down in service.  But solid state 
electronics (like a PCB) aren't usually exposed to conditions that 
result in wear-out, but there are exceptions (often due to thermal 
cycling or mechanical vibration).

There was a study conducted back in the 1960's by people in the 
Aerospace Industry to validate some of the assumptions about the nature 
of failures, and one of the major conclusions was that the "bathtub 
curve" that most people thought was so common actually occurs only a 
small fraction of the time (about 4% of all failures).  It actually 
turned out that infant mortality (without wear-out failures later in 
life as in a bathtub curve) actually accounted for about 68% of all 
failures.  The group that commissioned that study was called MSG-3 (if I 
remember correctly), and the results were discussed in a report called 
"Reliability-Centered Maintenance" (RCM for short) written Stanley F. 
Nowlan & Howard F. Heap.  RCM is now a standard basis for the planning 
of aircraft maintenance around the world, and is used in a wide variety 
of other industries as well.

I know this is a particularly specialized field and is pretty technical, 
but if anyone would be interested in learning more, I can provide more 
details and direct you to resources available on the Web.

>> So, without having specific data to refer to, I'd be inclined to say 
>> that 6 years isn't bad amount of service life to get out of that 
>> card, but the question of whether the other ones are about to fail is 
>> much more difficult to answer.
> Consider this -- for the last several generations, designs have been 
> getting generationally better roughly every 18 months, with the cycles 
> slowly getting shorter (so that we're getting closer to 12 month 
> generations).
>
> In that paradigm, the SCSI card that lasted six years equates to a 
> hardware device that has lived through four generations.  Now, when 
> you are the fourth generation in your family (and in your 80s, because 
> human generations are about twenty years), tell me how likely you are 
> to be working well?
>
> It's had a good life.  Time to give it a good burial and go buy a 
> replacement.
>
> If you want exactly the same device, or even something just 
> compatible, you're going to have a hard time finding it since most of 
> those will probably have died or been recycled already.
>
> You may even end up having to completely replace the entire computer 
> system in question, because you just can't find decent hardware that 
> is compatible with equipment that old.
I concur completely.

Al Lesmerises



More information about the SATLUG mailing list