Compiling ports leading to frying/overheating my system?

Hello, I seem to have a problem and I am not sure if FreeBSD is at fault or my hardware. Anyhow it seems that after putting my desktop system under a lot of stress by trying to compile many ports at once using synth that my hardware got damaged. What happens is that I will turn on my computer and it will go through the POST and BIOS checks, start to load FreeBSD and quickly turn off after 30 seconds or a minute or so. This desktop is something that I built myself a few years ago and I was having issues with it recently and replaced both the PSU and the motherboard. The desktop right now is out of commission and I have not been able to use it for a few weeks now. Right now I am using my laptop which has Debian loaded on it. I would like to put FreeBSD on the laptop but I don't want to put it on there if it is going to fry my system and kill my laptop as well.

What do you guys think? Can overstressing the system by compiling many things at once permanently damage my hardware? I would like to put FreeBSD on my laptop but I don't want to overstress the laptop and have that die as well.
 
Hello, I seem to have a problem and I am not sure if FreeBSD is at fault or my hardware.
Fire up a live media and test.

Anyhow it seems that after putting my desktop system under a lot of stress by trying to compile many ports at once using synth that my hardware got damaged.
The stress will cause the CPU overheating, and may be the RAM overheating too, hoverwer the CPUs are designed with some thermal protection. Worth to perform a RAM test.

What happens is that I will turn on my computer and it will go through the POST and BIOS checks
CMOS battery status ?
 
If your hardware is dusty, causing it to not be able to cool well. I've taken some of my hardware apart, some of which doesn't violate warranty, and dusted it off carefully (stay away from capacitors in the power supply [not really supposed to be tampered with], because they can hold a charge enough to shock, when disconnected), and reapplied thermal paste on the CPU. After that, my fan stopped making loud noise, and my computer stopped shutting off during compiles. A safety feature is to set it in bios to warn you, or to shut off when the CPU core reaches a certain temperature.
 
What happens is that I will turn on my computer and it will go through the POST and BIOS checks, start to load FreeBSD and quickly turn off after 30 seconds or a minute or so.
Then this has nothing to do with FreeBSD since your computer has not started loading any software yet or, perhaps, had only begun to do so. In any case, no, FreeBSD did nothing to your computer. Feel safe putting it on your laptop or any other properly put together desktop system you may have.

Your issue is clearly one of hardware failure of components or improper piecing together of the hardware if it truly is overheating. Others have pointed out problems with heat sinks not seated well.

I don't know how synth puts together ports in any more stressful way than portmaster or make would but I can't see overheating being caused by the OS or synth. Consider, too, that thousands of FreeBSD users do this sort of thing every day.
 
I don't know how synth puts together ports in any more stressful way than portmaster or make would but I can't see overheating being caused by the OS or synth. Consider, too, that thousands of FreeBSD users do this sort of thing every day.

synth is a parallel builder and portupgrade/portmaster are not. They built one port at a time at -jX while synth builds 1 port x #builders at -jX. It also puts a lot of stress on the sheer amount of mount/umounts going on. It's by far more stressful than anything except poudriere. Poudriere is likely more stressful than Synth.
 
I just turned on the desktop yesterday and it booted up and is working fine. I think the problem is that i have a four year old liquid cooling system for the CPU and that the CPU just overheated. I am going to order a new air based cooling system and then I'll give synth another try.
 
From a guy who tinkers with an engine every now and then, and once made my own 1500 watt generator, I can tell you liquid cooling is superior to air cooling. That is why liquid cooling was invented. Maybe there is something wrong with your liquid cooling system? Pump worn out and running slow? Liquid coolant somehow got contaminated? The liquid-to-air exchanger (radiator) is clogged with dust, or blocked by something; books, pair of jeans? Did a coolant hose kink some which is restricting flow?

If anyone has heated up their CPUs its me, and mine have had no issues. Mind you, I keep them clean, and that reminds me they are due for a cleaning. My machines are all air cooled.
 
If it is one of those closed systems, like from Corsair ones, they seem to often have pump failures. I think most of them are made by the same manufacturer.

I never used water cooling due to they all need more maintenance. The custom ones are better but more complicated to maintain. I prefer to simple stick with some big/huge Noctua - they are ugly but works. :)

However, if there is no pump issues, I would first just replace the thermal grease, four years may be enough to screw up some greases.
 
As people have mentioned, reseat your CPU. My take is that you may have put too much paste on there in the first place. It then will degenerate when it gets hot and you get what you have now. Use only the thinnest layer you can smear.
 
What I read somewhere, long ago---and what's always worked for me---is to place the tiniest drop on the center of the CPU die, then lower the cooler/heat sink straight down onto the die, apply a little pressure to the cooler/sink, and fix lock cooler/heat sink in place. This should evenly spread the paste and create the proper seal.

It's also good practice to clean out the case and replace the paste at least once a year. If it's a tower that sees a lot of CPU-intensive activity, once every six months might not be a bad idea. You can usually tell the paste needs replacing as soon as you remove the cooler, since it gets dry and flaky over time.
 
Currently there are those "liquid metal" ones from Coollaboratory, apparently they can be kept forever since they are 100% metal, but they should only be used (IIRC) on heatsinks with copper or nickel bases.
 
It's also good practice to clean out the case and replace the paste at least once a year.
FWIW, my "best practice" goes like this
cleaning dust (filters and case insides) - every 6 months
replacing thermal paste - every 3 years or when required (when temperatures start to go higher than normal over a period of several weeks)
my machines runs 24x7
YMMV.
 
Well I installed my new air based CPU cooler, used a small amount of arctic silver compound and installed it. It works fine until I use synth then it overheats and the computer shuts off within about 5 minutes. I am using the command synth upgrade-system. For some reason the heat from the CPU is not being distributed through the heat sink.
 
It works fine until I use synth then it overheats and the computer shuts off within about 5 minutes.
That means two things:
1) you are not using any CPU/RAM intensive software other than synth
2) your system do have some issue with heat / airflow or related things, like you seems to confirm:
For some reason the heat from the CPU is not being distributed through the heat sink.

However, unless you are already monitoring the CPU temperature and had proof of overheating, I would not exclude other components failures ...
 
However, unless you are already monitoring the CPU temperature and had proof of overheating, I would not exclude other components failures ...

Yep. I would verify CPU temp during a synth run: sysctl -a|grep temperature

Arctic silver is good stuff but it's possible the heat sink is not installed properly. If CPU temps are OK my next guess would be the power supply, then RAM, then general motherboard flakiness. The CPU itself being bad is possible, but very unlikely.
 
Back
Top