Musings on hardware prices - opps

Wed Aug 24 00:41:39 PDT 2005

On Aug 23, 2005, at 20:34, Michael T. Halligan wrote:

>> yup... seems to be par ... and i say one gets what one pays for ??
>>
>>
>>
> Yes.  Now the math here works out that
> 1000 servers from Dell cost $1M
> 1000 HP servers cost $2.M
> 1000 IBM servers cost $2.8M
>
> If you buy those 1000 servers, using the 30% rule, that means you  
> only "needed" 769 servers, and
> over-bought those 231 servers for redundancy, and extra performance  
> just in case.

All that has me thinking (again) about the hidden costs of hardware  
on the cheap.  Let's look again at those figures for the 1000  
servers.  If you're saying that Dell's stuff fails 30% more often  
than HW from HP, IBM, or SUN, then you really should be comparing  
1000 servers from Dell with 700 servers from the "premium" players  
(where'd you get 70% of 1000 = 769?).  So that'd be

1000 servers from Dell cost $1M
700 HP servers cost $1.4M
700 IBM servers cost $1.96M

Still, the HP boxes cost 40% more than those from Dell, and IBM's are  
about twice as much as Dell's, but things aren't quite so heavily  
weighted as you made them seem, Michael.

[...snip...]

>>> As much as I despise
>>> Dell's component choices and technical support, it's a financial  
>>> AND technical decision dell is making really hard.
>>
>> yup...if one is interested in pricing ... dell is hard to beat
>>
>> if the customer is intested in performace/reliability ... than  
>> dell is out of the loop in my book
>>
>>
> Blanket statements are really hard to make on imperial evidence. My  
> rule of thumb is
> to expect Dell gear to be down 30% more often than HP or IBM.  
> That's to build in room for
> problems. In actual experience, I can look over uptime reports for  
> two customers in 2004. One
> of them had 100 Dell servers. One of them had 45 HP servers. The  
> dell environment, overbuilt,
> had 99.99% uptime. The HP environment, if I adjust the #s, had  
> about 99.95% uptime.  Service
> uptime during a 12 hour business day, both archived 5 9s.
>
> The difference? The Dell environment was more of a pain to install  
> initially, and had more server failures.

Ah, now there's the rub:  what do you do with those 300 extra (and  
now dead) Dell boxes that you had to buy due to the higher failure  
rate of Dell products?  You get 'em fixed/replaced, of course--Dell's  
great at it with their service tag thingy!  OK, but _someone_ still  
has to make all the phone calls (hey, they don't all fail at once),  
pack the returns, unpack the replacement parts, install 'em, perhaps  
rebuild entire boxes, and so forth.  Ick!

It doesn't stop there, either, because 30% of the replacement  
equipment also will fail, nullifying 100 of the 300 needing repair  
and costing even more in time/effort--and, of course, 30 of those 100  
will break, and 10 of those 30, which may get you down to the failure  
rate of the HP gear (don't know--haven't ever been in a position to  
work with such numbers).

The thing is that if, as is my current situation, your colo is 30  
minutes away without traffic, you're talking about three hours at  
least per repair/replacement, factoring in phone, travel, part  
receipt, replacement, & return times.  The average to-fix time could  
be significantly higher, depending on the particular HW involved &  
how it failed (eg. disks are easy if things are mirrored, but you  
reinstall everything if not).  As you see, this can significantly  
increase the cost of going with Dell boxen (say we need to fix those  
300 + 100 = 400 systems, at an average repair time of 4 hrs each, and  
sysadmin time costs $100 / hr--all somewhat mushy figures, but you  
have to pick _something_).

As if all that wasn't hard enough to quantify, there's the morale of  
dealing with failing HW.  You said, "hardware sucks," Michael, and  
that you preferred to essentially ignore the HW & concentrate on the  
infrastructure.  I agree, mostly:  from my point of view, HW is part  
of that infrastructure, and doing a job as well as possible to me  
means using HW that doesn't suck.  I like to see well-designed HW as  
much as I like to see well-designed software or a well-designed  
architecture.  When things suck, sysadmins move elsewhere; when they  
don't, we stay put.  Training new sysadmins in an environment adds to  
the cost of purchasing crappy HW.  It's not something that's easy to  
measure, but it exists nonetheless.

-Guy