Quieting down HP Proliant DL380 G5 server for home use

mh321

Member
Jul 16, 2013
33
0
0
Hello,
I recently got the above server with a bunch of other components for free. I got the server to run windows server 2003 without any problems. It has one 2.83ghz Xeon quad core CPU, so it is by far the fastest computer I have in the house, and I'm hoping on keeping it as the build quality is nice. I want to try to somehow replace my P4 PC with it.

How can I quiet it down without spending a lot of money? The vast majority of the noise comes from the fans. All the fans pull out easily out of their special slots, what can I replace them with? I know that quiet desktop PC fans are cheap, will they work?

The server has 6 fans (excluding the psu fans), and they are all:

Nidec Beta V
TA225DC
Model B35441-94
12V DC 1.5A
60mm size
They have 4 wires but a 2x3 connector



I'm adding photos of the server to better the chance of getting some help with the fans.







Fans removed:





Thanks
 

caution

Member
Jul 16, 2013
37
0
0
Server's aren't build with quiet operation in mind, indeed.

I wouldn't bother with finding silent 60mm fans. Get some nice 120mm (BeQuiet/NB/Noctua/etc) and go ghetto on that puppy
 

mh321

Member
Jul 16, 2013
33
0
0
Any ideas on how to get bigger fans in, while still being able to close the cover? At an angle maybe? The stock fans have 4 wires in them, which from looking at other fans on sale, is quite unusual, are all 4 wires needed?
 

imagoon

Diamond Member
Feb 19, 2003
5,199
0
0
There is no easy way to use that as a desktop. It is a rack mount case that expects a certain amount of airflow and certain RPM fans. There is not going to be a way to install larger fans that won't trigger a BIOS shutdown without spending more money than it would cost to get a Desktop from the Dell outlet. I mean you can get refurbished desktops off the Dell site for $99 -> $149 that are going to be pretty comparable to that server. That is a G5 which is very out of date. Heck the G6 series is basically gone now also.
 

mh321

Member
Jul 16, 2013
33
0
0
So the server is able to detect the fan rpm using the 4 cables? And if it doesn't read the rpms it was hard-coded to expect, it will get confused and shut down?

Does this server use considerably more power than a desktop, or the same?

Any good ways to build a sound absorbing case for it, and put it in the closet? I've got some insulation foam, plexiglass, and plywood at home
 

mlody

Senior member
Apr 10, 2001
277
0
76
if you are allowed, I would resell it and use that money to upgrade your main computer (if possible) or build a new computer.
 

imagoon

Diamond Member
Feb 19, 2003
5,199
0
0
So the server is able to detect the fan rpm using the 4 cables? And if it doesn't read the rpms it was hard-coded to expect, it will get confused and shut down?

Does this server use considerably more power than a desktop, or the same?

Any good ways to build a sound absorbing case for it, and put it in the closet? I've got some insulation foam, plexiglass, and plywood at home

Yes it reads the RPM from the 4 wires. G5's should shut down if they detects 2 fan failures in a bank. I am pretty sure they are not standard fans (in the home computer world at least) I am pretty sure they normally have 12volts at the fan all the time and use one wire to send a PWM signal to control speeds and the other to detect RPM. This is closer to the 4 and 5 wire Intel CPU fans.
 

mh321

Member
Jul 16, 2013
33
0
0
How is the power consumption of this server compared to that of a desktop PC, if I was able to get around the noise problem by building a special cabinet for it?

Does the server use a "scary" amount of electricity?

Would it make a big impact on my electricity bill if it ran 4 hours a day?

Thanks for everyone's advice so far
 

imagoon

Diamond Member
Feb 19, 2003
5,199
0
0
How is the power consumption of this server compared to that of a desktop PC, if I was able to get around the noise problem by building a special cabinet for it?

Does the server use a "scary" amount of electricity?

Would it make a big impact on my electricity bill if it ran 4 hours a day?

Thanks for everyone's advice so far

As compared to?

A server part for part compared to a desktop will be less efficient due to multiple power supplies. However most servers are multiple CPU, have SAS / SCSI cards, much faster disk RPMs where desktops do not. This normally means that they can pull several hundred more watts than a desktop.
 

mh321

Member
Jul 16, 2013
33
0
0
...compared to an average store-bought PC. Or my current almost decade-old hp pavilion a824n PC (has a 300W PSU if that helps). My PC has 2x 7200rpm HDDs and a FX5500 GPU.


This server is single CPU, one SCSI card, and has 2x 10k serial scsi HDDs. It does have 2 power supplies though, which I want to ask also if it uses more power if both PSUs are plugged in?

With this info, would there be a big difference in electricity? 90% of usage would be just for low CPU intensity tasks like internet.
 

imagoon

Diamond Member
Feb 19, 2003
5,199
0
0
Well those tend to come with dual 750W or Dual 1000W power supplies. That era also may not support any power saving modes. The machine will complain and run the fans at a higher speed if only one PSU is plugged in.

This machine isn't a desktop PC. Is there some reason you plan to spend more money to make it one then just buying a desktop PC?

I have a 380G6 sitting here and it has dual 750watt supplies and it is pulling a cool 560 watts chilling on the desk.
 

mh321

Member
Jul 16, 2013
33
0
0
Well it was free, and if I sold it, I probably would just save the money, as I don't -have- to have a new computer since mine still runs. I just see this as an opputunity to have something newer. I guess I also like the build quality and it's kind of cool.

I kind of solved the noise problem, I tried running it in my closet, and found out it's not noisy at all in there. There are ventilation holes in the closet too. If I built a cheap cabinet for it, it should barely be audible.

I had plans on adding a nice $20 sound card to it, along with a radeon 7750 for $90. I still have to decide, but $110 does not seem too bad and both parts can always be transfered to another computer.

The only thing that would worry me is power consumption, that's 560W just with the windows desktop on and no other programs? Is that like having 2 computers running at the same time?
 

mlody

Senior member
Apr 10, 2001
277
0
76
The only thing that would worry me is power consumption, that's 560W just with the windows desktop on and no other programs? Is that like having 2 computers running at the same time?

I think 560w continuous use is more like 10 computers with light loads. My new i7-4770 with on board video, 4 dimms, 2x hdd, 1x ssd idles in the desktop using about 30-40w according to the readings of my APC ups - monitor off.
If my cpu is running handbrake workloads at %100 i see my power consumption going over 100watts.
 

imagoon

Diamond Member
Feb 19, 2003
5,199
0
0
Most servers of that generation didn't really have the power savings modes that the new gear has. This box pulling 560 is pretty much idle at the moment. Sub 5% CPU, not a lot of disk IO etc.
 

mh321

Member
Jul 16, 2013
33
0
0
I'll see if I can sell it then. Any good places to sell servers other than craigslist or ebay?
Thanks
 

Chris Wager

Junior Member
Mar 11, 2017
6
0
1
Okay so this is an old thread but if I found it others might.

The DL380 G5 is far from almost gone now and the G6 certainly isn't. The G5 officially supports Server 2008 R2 and is still in use widley. And is also popular in esxi environments for virtualisation still used by at&t. The G5 consumes about 315W on average. The average gaming pc cobsumes about 350w to 450w they are comparable but the G5 is designed to compute processes and network resources faster than an average of cpu.

If you want a quite home server try a Fujitsu Siemens Primergy M series they are tower form factor and run about as quite as a PC unless extremely hot or under extreme load.

That being said I use a dl380 g5 as a home server and do not find it that noisy. If yours is excessively noisy replace the power supplies they are likely faulty. Was a common fault on these rigs.

The other option if course of you don't need all that extra horse power and only nerd 1 Windows server would be to by a Qnap Nas which supports a 4 disk sata array. Up to 8gb of ram and has a hypervisor built in to run Windows server and that will run almost silently and use very little power.
 

Tarheelbandb

Junior Member
Apr 18, 2017
3
0
1
Okay so this is an old thread but if I found it others might.

The DL380 G5 is far from almost gone now and the G6 certainly isn't. The G5 officially supports Server 2008 R2 and is still in use widley. And is also popular in esxi environments for virtualisation still used by at&t. The G5 consumes about 315W on average. The average gaming pc cobsumes about 350w to 450w they are comparable but the G5 is designed to compute processes and network resources faster than an average of cpu.

If you want a quite home server try a Fujitsu Siemens Primergy M series they are tower form factor and run about as quite as a PC unless extremely hot or under extreme load.

That being said I use a dl380 g5 as a home server and do not find it that noisy. If yours is excessively noisy replace the power supplies they are likely faulty. Was a common fault on these rigs.

The other option if course of you don't need all that extra horse power and only nerd 1 Windows server would be to by a Qnap Nas which supports a 4 disk sata array. Up to 8gb of ram and has a hypervisor built in to run Windows server and that will run almost silently and use very little power.

Thanks for digging up this old thread. Most people like to hate on it but I love it.
I just picked one up locally off ebay for $20.50. It's running (2) 2.33GHZ but I just found the 3.33GHZ (x5470) on ebay for $24 each in case you are interested. That would make the G5 (back then) twice as good as the i7-995. I just placed the order today. And when I say on par I mean that without exaggeration. Just check out the comparison. http://cpuboss.com/cpus/Intel-Xeon-X5470-vs-Intel-Core-i7-975 I'm currently running Xenserver, trying to get some proficiency at running VMs.

I also don't really understand the earlier comments about being able to purchase a better PC, especially at the time the comments were written. In terms of power usage the X5470 use half the power the I7 does meaning it costs around the same per year to run the dual chip set up on the G5.

I came here looking for an issue with the fan noise. I am assuming the issue is the power supply as I can measure the fan output of the PSU at about 20 dB higher than the front. I just ordered a new PSU for $14 bucks as well
 

XavierMace

Diamond Member
Apr 20, 2013
4,307
450
126
So much incorrect info at once, it's hard to know where to start.

A) Adding a second processor doesn't automatically make it twice as fast. An i7-975 is faster in single threaded operations, adding a second X5470 doesn't change that.
B) I don't know where you came up with the idea that the X5470 uses half the power, but that's not remotely true. The X5470 has a TDP of 120W, the i7-975 a TDP of 130W but the X5470 is a generation older and C states keep getting better and better. A pair of X5470's is going to be pretty close to double the power consumption of a i7-975
C) Even at the time this thread started, the G6's were already dirt cheap and use far less power.
 

Tarheelbandb

Junior Member
Apr 18, 2017
3
0
1
First I'd love to see your sources, because I am basing my info on the link I provided.

A. Sure not automatically but in fact this case it does. 1. Because the X5470 has more cache (12x). That is a metric. 2. Windows Server 2008 (the latest server running well on the server) takes advantage of all 8 cores available. 3. The single i7 per CPU boss marginally edges out the X5470. 4. They both have basically the same score. I'll Again..I do not mind if you would cite your sources.

B. Once again CPUboss states what I repeated. Half the power and the OP was running a single chip. Sure c states are getting better and better but the power hungry i7 was still power hungry. All I know is that for the last 24 hours mine has averaged 303 watts peaking at 374. Seems on par with my statement *Kanye Shrug*

C. Per this post (https://serverfault.com/questions/243439/hp-proliant-dl380-g4-can-this-server-still-perform-in-2011 in 2011) , G5s were still going for over $700 on a "good deal". I remember them still being close to $1k even though they were retired by HP in 2009. If you consider that dirt cheap then kudos to you. Sure relative to newer ones, but hardly qualifies as "dirt cheap". A quick eBay search shows them being sold regularly for around $$100- 300. That's for a decade old piece of equipment. So I would love to see your es to support this dirt cheap claim.

I am not afraid to be corrected. Especially if you just flat out say the basis of my assumptions (CPUboss) is wrong. I don't own both to know myself.
 
Last edited:

Chris Wager

Junior Member
Mar 11, 2017
6
0
1
@Tarheelbandb The chances are you power supplies are the cause of the loud noise, I am currently working a datacentre with 28u racks each holding 8 of theses and about half of them are suffering loud noise from fans, based on audit I have confirmed all the ones that are not affected have had recent power supply replacements. if you replace on you will need to replace both because otherwise you will only place excessive strain on the new power-supply due to the ageing components in the remaining old supply, also i would not recommend running this on a single psu either in case you are considering doing so the insight controller will see that as a fault and increase the fan speeds to keep the remaining power supply cool again making noise excessive. Your CPU comparison is fairly accurate, the i7 is a desktop cpu the haters wont like you bench marking that against an older xeon chip but I agree with and support you comparison and given the fact I have some old rack server that use Pentium 4 CPU's and an almost new single processor core i7 rack server I would say that was a very accurate and fair test and comparison actually and @XavierMace comment that the G6 was already dirt cheap and used far less power is laughable the G6 uses almost identical hardware to the g5 so much so that they share almost all the same driver I have a second g5 running server 2012 using all the drivers and smart start from the G6 and it is more stable than my other g5 using the orginal g5 driver on 2008 shows just how the same the hardware is, and I have a g6 in a work lab running nothing but iis, and it uses average 340-358 watts just running a simple html page on iss and I see the same preformance on other G6 rigs at idle my G5 uses 313 average under nominal load active directory and asterisk in a hyper-v vm so that suggest that the G5 actually uses less power, I have E5440 CPU's at 2.8Ghz each is a quad core and 32GB ram in my g5 2012 setup and the same in my 2008 rig both average the same power consumption under similar loads
 
Last edited:

Chris Wager

Junior Member
Mar 11, 2017
6
0
1
First I'd love to see your sources, because I am basing my info on the link I provided.

A. Sure not automatically but in fact this case it does. 1. Because the X5470 has more cache (12x). That is a metric. 2. Windows Server 2008 (the latest server running well on the server) takes advantage of all 8 cores available. 3. The single i7 per CPU boss marginally edges out the X5470. 4. They both have basically the same score. I'll Again..I do not mind if you would cite your sources.

B. Once again CPUboss states what I repeated. Half the power and the OP was running a single chip. Sure c states are getting better and better but the power hungry i7 was still power hungry. All I know is that for the last 24 hours mine has averaged 303 watts peaking at 374. Seems on par with my statement *Kanye Shrug*

C. Per this post (https://serverfault.com/questions/243439/hp-proliant-dl380-g4-can-this-server-still-perform-in-2011 in 2011) , G5s were still going for over $700 on a "good deal". I remember them still being close to $1k even though they were retired by HP in 2009. If you consider that dirt cheap then kudos to you. Sure relative to newer ones, but hardly qualifies as "dirt cheap". A quick eBay search shows them being sold regularly for around $$100- 300. That's for a decade old piece of equipment. So I would love to see your es to support this dirt cheap claim.

I am not afraid to be corrected. Especially if you just flat out say the basis of my assumptions (CPUboss) is wrong. I don't own both to know myself.

@Tarheelbandb The chances are you power supplies are the cause of the loud noise, I am currently working a datacentre with 28u racks each holding 8 of theses and about half of them are suffering loud noise from fans, based on audit I have confirmed all the ones that are not affected have had recent power supply replacements. if you replace on you will need to replace both because otherwise you will only place excessive strain on the new power-supply due to the ageing components in the remaining old supply, also i would not recommend running this on a single psu either in case you are considering doing so the insight controller will see that as a fault and increase the fan speeds to keep the remaining power supply cool again making noise excessive. Your CPU comparison is fairly accurate, the i7 is a desktop cpu the haters wont like you bench marking that against an older xeon chip but I agree with and support you comparison and given the fact I have some old rack server that use Pentium 4 CPU's and an almost new single processor core i7 rack server I would say that was a very accurate and fair test and comparison actually and @XavierMace comment that the G6 was already dirt cheap and used far less power is laughable the G6 uses almost identical hardware to the g5 so much so that they share almost all the same driver I have a second g5 running server 2012 using all the drivers and smart start from the G6 and it is more stable than my other g5 using the orginal g5 driver on 2008 shows just how the same the hardware is, and I have a g6 in a work lab running nothing but iis, and it uses average 340-358 watts just running a simple html page on iss and I see the same preformance on other G6 rigs at idle my G5 uses 313 average under nominal load active directory and asterisk in a hyper-v vm so that suggest that the G5 actually uses less power, I have E5440 CPU's at 2.8Ghz each is a quad core and 32GB ram in my g5 2012 setup and the same in my 2008 rig both average the same power consumption under similar loads
 

Tarheelbandb

Junior Member
Apr 18, 2017
3
0
1
@Chris Wager thanks for that info. I hadn't considered actually using both PSUs for load balancing/noise reduction. I just assumed one did not work untill the other failed via some sort of relay. I've ordered a 2nd 1000W PSU per your suggestion.
 

Chris Wager

Junior Member
Mar 11, 2017
6
0
1
@Chris Wager thanks for that info. I hadn't considered actually using both PSUs for load balancing/noise reduction. I just assumed one did not work untill the other failed via some sort of relay. I've ordered a 2nd 1000W PSU per your suggestion.
The dl380 load balances the power usage across both powersupplies and continues to operate on one if the one power supplies fails, so its best for both noise, temp and performance to use both power supplies, the dynamic power regulation also needs both power supplies to work effectivley.
 

XavierMace

Diamond Member
Apr 20, 2013
4,307
450
126
First I'd love to see your sources, because I am basing my info on the link I provided.

A. Sure not automatically but in fact this case it does. 1. Because the X5470 has more cache (12x). That is a metric. 2. Windows Server 2008 (the latest server running well on the server) takes advantage of all 8 cores available. 3. The single i7 per CPU boss marginally edges out the X5470. 4. They both have basically the same score. I'll Again..I do not mind if you would cite your sources.

B. Once again CPUboss states what I repeated. Half the power and the OP was running a single chip. Sure c states are getting better and better but the power hungry i7 was still power hungry. All I know is that for the last 24 hours mine has averaged 303 watts peaking at 374. Seems on par with my statement *Kanye Shrug*

C. Per this post (https://serverfault.com/questions/243439/hp-proliant-dl380-g4-can-this-server-still-perform-in-2011 in 2011) , G5s were still going for over $700 on a "good deal". I remember them still being close to $1k even though they were retired by HP in 2009. If you consider that dirt cheap then kudos to you. Sure relative to newer ones, but hardly qualifies as "dirt cheap". A quick eBay search shows them being sold regularly for around $$100- 300. That's for a decade old piece of equipment. So I would love to see your es to support this dirt cheap claim.

I am not afraid to be corrected. Especially if you just flat out say the basis of my assumptions (CPUboss) is wrong. I don't own both to know myself.

A. CPUBoss is wrong. As stated in my previous post, the 975 is a 130W TDP (http://ark.intel.com/products/37153/Intel-Core-i7-975-Processor-Extreme-Edition-8M-Cache-3_33-GHz-6_40-GTs-Intel-QPI?q=i7 975). I don't know where the hell CPU boss is coming up with 212W. Certainly not at stock clocks. The cache is also wrong, the 975 is a 8MB cache and faster. The X5400's don't have hyperthreading or turbo. Comparing (on CPUBoss) the X5470 to the X5460 (same processor, 200mhz slower clock), they state the X5470 is 50% faster on the Geekbench test. Bullshit. So, I wouldn't trust their numbers. Stating "Windows Server 2008 takes advantage of all 8 cores" would seem to indicate to me that you're not aware of how multithreading works. Tossing more threads at Windows doesn't magically make it faster.

B. 303W is a LOT. For comparison, my DL380G6's with 2x L5640's (6C/12T) and 128Gb of RAM pulls 170W under my normal operating load (half dozen VM's).

C. I'll give you the price because I don't recall what I paid back then. $150 is normal these days.

@Tarheelbandb @XavierMace comment that the G6 was already dirt cheap and used far less power is laughable the G6 uses almost identical hardware to the g5 so much so that they share almost all the same driver I have a second g5 running server 2012 using all the drivers and smart start from the G6 and it is more stable than my other g5 using the orginal g5 driver on 2008 shows just how the same the hardware is, and I have a g6 in a work lab running nothing but iis, and it uses average 340-358 watts just running a simple html page on iss and I see the same preformance on other G6 rigs at idle my G5 uses 313 average under nominal load active directory and asterisk in a hyper-v vm so that suggest that the G5 actually uses less power, I have E5440 CPU's at 2.8Ghz each is a quad core and 32GB ram in my g5 2012 setup and the same in my 2008 rig both average the same power consumption under similar loads

You're on crack. The G6 and G7's are almost identical, the G5 and G6's aren't remotely close. They "share the same driver" to the extent Intel's driver package covers pretty much every chipset they've made in the last 15 years.

G5: Intel 5000 chipset, LGA771 Woorcrest-Harpertown processors (4C/4T Max), FB-DDR2 RAM.
G6: Intel 5520 chipset, LGA1366 Nehalem EP-Westmere EP processors (6C/12T Max), DDR3 RAM.

The dl380 load balances the power usage across both powersupplies and continues to operate on one if the one power supplies fails, so its best for both noise, temp and performance to use both power supplies, the dynamic power regulation also needs both power supplies to work effectivley.

IF you've set them to balanced in the BIOS, then yes they will be balanced. Given the age of these systems, I'll give you that you MAY be better off running both. On newer systems, that's less likely. Power supplies are generally at peak efficiency around 80-90% load. Running two power supplies at 40% load is likely using more power than one power supply at 80% load. Dynamic Power Regulation is purely controlling the C states of the processors and is not at all effected by power supply count.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |