[AccessD] disk drive temps for the Norco 4020

jwcolby jwcolby at colbyconsulting.com
Tue Jul 20 14:26:19 CDT 2010


You laugh... I have considered moving it to the basement.  The basement is pretty cool all year, and 
very cool in the winter.  Once I get everything migrated to rack mount I could easily drop a little 
window AC into one of the windows in the basement, and then build a little closet to contain the 
cold air.  Good idea!

John W. Colby
www.ColbyConsulting.com


Michael Bahr wrote:
> John, the water cooling is intriguing.  I researched it several years ago
> when I over-clocked my cpu.  I found information on Tom's Hardware that
> compared water-cooling to air cooling.  The water-cooling did lower the
> cpu temp but not as much as they had hoped once the system stabilized. 
> Comparing this to a **high quality** air cooler the differential was
> **not** that great.  Again the water-cooler is subject to the ambient
> temps.  Considering the water-coolers have thier own hazards like leakage,
> pump failure, coolant replacement.  Some water-coolers circumvent some of
> the hazards by being a permanent closed system.  But then...   Actually
> the ultimate cpu cooler uses a refrigerant-type coolant just like your
> refrigerator.
> 
> But the main goal I was referring to was not just the cpu but the whole
> system.  The HD temps need to be controlled.
> 
> I suggest make your system portable.  Move it to your living room during
> the winter (to save on heating costs) and then move it to the basement
> during the summer.  :-)
> 
> Mike...
> 
>> Michael,
>>
>> Yup, yup and yup.  Yes, of course the lower the ambient (starting point)
>> the lower all temps from
>> that point.  However I really cannot reasonably get the system into a
>> controlled (very cold)
>> environment.
>>
>> As far as the CPU temps go, the best strategy would be water cooling,
>> simply because then the
>> ambient at the surface of the CPU can be controlled directly by the temp
>> of the water, and also
>> because you get high thermal mass fluid removing the heat which is dumped
>> outside of the case.
>>
>> I am moving to a rack mount system.  Once that move is complete, I may
>> look at a single water
>> cooling system for all my servers.  For the moment, 130F isn't extreme
>> enough to spend time on it.
>> I will be keeping an eye on it though.
>>
>> John W. Colby
>> www.ColbyConsulting.com
>>
>>
>> Michael Bahr wrote:
>>> John, the temps are largely dependent on the evironment.  Having your
>>> system in a temperature-controlled room like the host servers do will
>>> lower the temps.  Those environment are usually around 60 degrees F.  If
>>> you have your system tucked away in your house and the temp is ~75
>>> degrees
>>> F, then you **may be** pushing the limits.
>>>
>>> You really need to find some alternate means to lower the temps.  One
>>> easy
>>> solution is to find a high velocity fan(s) that you can replace in your
>>> system.  It will make some noise but better fan noise than HD
>>> noise--:-O.
>>> Maybe a high performance CPU cooler like one from Zalman.
>>>
>>> With my over-clocked system I run a Zalman cooler and a high-velocity
>>> fan
>>> which reduced my CPU temps by amost 10 degree C.  When I process videos
>>> my
>>> CPU jumps up from ~41C to ~55C and stays there for hours on end.  I
>>> believe the thermal shutdown of Intel CPU's is around 80 degrees C.  The
>>> cooler I can keep things the longer the electronics will last.
>>>
>>> Mike...
>>>
>>>> Michael,
>>>>
>>>> You are right it does look high, and no, I was not particularly
>>>> stressing
>>>> the system.  The problem
>>>> here is that I have 20 hard disks dumping heat into the back part of
>>>> the
>>>> box.  So the CPU is not
>>>> starting at ambient, it is starting at ambient plus disk heat load.
>>>>
>>>> To be honest I do not worry about the CPU until it gets well north of
>>>> 60C.
>>>>  In one of my many
>>>> previous lives I was tasked with running an oven that did a burn in on
>>>> PC
>>>> boards.  We burned in our
>>>> boards at 180 F (80C).  Of course those were mil spec and they were not
>>>> expected to run at that temp
>>>> forever but that did convince me that electronics can easily survive
>>>> the
>>>> occasional forays into
>>>> scorching temps.
>>>>
>>>> The real issue will come when I get a pair of 8 core or 12 core
>>>> processors
>>>> in the case.  These new
>>>> procs are rated to run all cores full tilt at under 125 watts but that
>>>> does add even more heat load
>>>> into the case.
>>>>
>>>> John W. Colby
>>>> www.ColbyConsulting.com
>>>>
>>>>
>>>> Michael Bahr wrote:
>>>>> John, the CPU temp looks unusually high, 55C.  Were you playing any
>>>>> games
>>>>> during the testing?  Is this the idle temp?
>>>>>
>>>>> Mike...
>>>>>
>>>>>> I bought the Norco 4020 to build a server with.
>>>>>>
>>>>>> http://www.newegg.com/Product/Product.aspx?Item=N82E16811219021&cm_re=4u_rackmount_case-_-11-219-021-_-Product
>>>>>>
>>>>>> I am impressed with the case, though it is consumer grade.  It is
>>>>>> tough
>>>>>> to
>>>>>> work with however.  It
>>>>>> has a fan wall just behind the disk drive cage which leaves exactly
>>>>>> two
>>>>>> inches between the fans and
>>>>>> the disk backplanes.  Obviously it is impossible to leave the wall in
>>>>>> place and get all of the
>>>>>> cables in, but removing the fan wall is a job for a safe cracker.
>>>>>>
>>>>>> None the less I did so, got 20 disks in and cabled up, then
>>>>>> reinserted
>>>>>> the
>>>>>> fan wall, and then
>>>>>> threaded the cables through the holes in the BOTTOM of the fan wall.
>>>>>> Why
>>>>>> on earth they didn't leave
>>>>>> the holes on the top of the fan wall is beyond me but they didn't.
>>>>>>
>>>>>> Anyway, for the first time tonight I tested temps with all 20 drives
>>>>>> inserted and power applied, and
>>>>>> the fan wall operating.  I ran a defrag on every volume
>>>>>> simultaneously
>>>>>> in
>>>>>> order to cause the disks
>>>>>> to work and generate as much heat as I could cause.
>>>>>>
>>>>>> Below are the temps.
>>>>>>
>>>>>> CPU Temperature 	55 ºC
>>>>>> Ctrl Temperature 	49 ºC
>>>>>> Power +12V 	12.099 V
>>>>>> Power +5V 	5.053 V
>>>>>> Power +3.3V 	3.328 V
>>>>>> SATA PHY +2.5V 	2.528 V
>>>>>> DDR-II +1.8V 	1.872 V
>>>>>> PCI-E +1.8V 	1.872 V
>>>>>> CPU +1.8V 	1.856 V
>>>>>> CPU +1.2V 	1.232 V
>>>>>> DDR-II +0.9V 	0.928 V
>>>>>> Hdd#1 Temperature 	37 ºC
>>>>>> Hdd#2 Temperature 	40 ºC
>>>>>> Hdd#3 Temperature 	42 ºC
>>>>>> Hdd#4 Temperature 	39 ºC
>>>>>> Hdd#5 Temperature 	47 ºC
>>>>>> Hdd#6 Temperature 	51 ºC
>>>>>> Hdd#7 Temperature 	49 ºC
>>>>>> Hdd#8 Temperature 	46 ºC
>>>>>> Hdd#9 Temperature 	37 ºC
>>>>>> Hdd#10 Temperature 	49 ºC
>>>>>> Hdd#11 Temperature 	44 ºC
>>>>>> Hdd#12 Temperature 	47 ºC
>>>>>> Hdd#13 Temperature 	45 ºC
>>>>>> Hdd#14 Temperature 	50 ºC
>>>>>> Hdd#15 Temperature 	49 ºC
>>>>>> Hdd#16 Temperature 	42 ºC
>>>>>>
>>>>>> There are only 16 drives showing because these 16 are cabled up to my
>>>>>> Areca raid co-processor card
>>>>>> which is a 16 port card, and that card provides a utility to watch
>>>>>> these
>>>>>> numbers.  The other disks
>>>>>> are not being driven but do have power applied.
>>>>>>
>>>>>> I am satisfied with the temps.  Several disks actually hit 53 degrees
>>>>>> C
>>>>>> under max stress but that is
>>>>>> acceptable to me.  I would like lower temps but I can live with
>>>>>> these.
>>>>>>
>>>>>> BTW I am using the Corsair 750W modular PS which is a NICE PS.
>>>>>>
>>>>>> http://www.newegg.com/Product/Product.aspx?Item=N82E16817139010
>>>>>>
>>>>>> The 12V cables are ribbon cables which makes them very easy to run
>>>>>> and
>>>>>> minimizes the air flow
>>>>>> impedance.  With only 2 inches to work with between the fans and
>>>>>> disks,
>>>>>> I
>>>>>> need all the help I can get.
>>>>>>
>>>>>> --
>>>>>> John W. Colby
>>>>>> www.ColbyConsulting.com
>>>>>> --
>>>>>> AccessD mailing list
>>>>>> AccessD at databaseadvisors.com
>>>>>> http://databaseadvisors.com/mailman/listinfo/accessd
>>>>>> Website: http://www.databaseadvisors.com
>>>>>>
>>>> --
>>>> AccessD mailing list
>>>> AccessD at databaseadvisors.com
>>>> http://databaseadvisors.com/mailman/listinfo/accessd
>>>> Website: http://www.databaseadvisors.com
>>>>
>>>
>> --
>> AccessD mailing list
>> AccessD at databaseadvisors.com
>> http://databaseadvisors.com/mailman/listinfo/accessd
>> Website: http://www.databaseadvisors.com
>>
> 
> 



More information about the AccessD mailing list