Re: Weakest point of a server?
On Thu, 6 Feb 2003 21:14, Jason Lim wrote:
> We know sleeve-bearing fans die pretty quickly and that ball-bearing fans
> tend to keep running much better/longer, but do you know approximately how
> long we're talking about? Is "long" 3 years, 5 years, or possibly even
> longer than that?
I'm working with 30+ machines that are 3 years old that (AFAIK) have not had a
failure on any of their fans. As each machine has a minimum of 1 CPU and and
2 case fans and as some machines have drive arrays with several fans the
total must be at least 200 fans. 200 fans and 3 years without a failure
suggests that they last well.
> How long should one go before replacing a ball-bearing fan?
I'd rather have a good way of detecting a dieing fan to replace it.
> can have redundancy (RAID 1,5,etc.) but I've rarely heard people talk of
> "redundant fans". Even in expensive DELL and HP servers there are usually
> only 1 fan on each CPU.
The Dell2650 machines I'm working with now are supposed to continue working
after one fan dies.
> > Software RAID can deal with this.
> I will investigate this further.
It's not nice but it works. Once you have replaced both disks with bigger
ones you can re-create a RAID-1 with a larger size. The same should be
doable with RAID-5 with much greater difficulty.
> So I guess it is safe to say that in genreal, that CPU and RAM will not
> fail provided they are not tampered with.
Also they have to be installed correctly. I have heard accounts of
incorrectly installed fans causing a CPU failure many months later.
> Temperature is usually kept constant. Depending on where the server is in
> the rack (top/bottom) the temperature tends to vary (top tends to be
> between 2-3 degrees higher than bottom). On that note, what temperature
> difference do you observe between top/bottom?
I have never measured it. Ideally the difference should be small enough that
you can't feel it with your hand. Sometimes the bottom server feels cold and
the top server is too hot to comfortably touch!
> The thing that really bites is that "40Gb" hard disks from different
> manuacturers seem to have quite different formatted capacities... heck,
> we've seen different capacities from the same manufacturer but slightly
> different model numbers (but the same model)!
Sure. Manufacturers often change their process to produce a cheaper part with
the same specs. When a certain size of drive is new it might take 3 platters
to store that capacity, when technology advances they make drives that store
in in 2 platters for a slight variation in capacity. It's no big deal.
See http://www.coker.com.au/~russell/hardware/46g.png for a graph of
performance in the Y axis and position on the disk in the X axis and then you
probably won't want to use all the disk anyway. ;)
> I guess one way would be to pre-purchase a whole bunch of matching-size
> drives, but then you run the risk of only using them a couple of years
> later, and then they might not start up at that time :-/
The main issue IMHO is that in a couple of years time you will be able to buy
larger and faster drives for less cost!
http://www.coker.com.au/selinux/ My NSA Security Enhanced Linux packages
http://www.coker.com.au/bonnie++/ Bonnie++ hard drive benchmark
http://www.coker.com.au/postal/ Postal SMTP/POP benchmark
http://www.coker.com.au/~russell/ My home page