Random, random freezes on several distros

[JayM]

IME random crashes/freezes are most often caused by:

Too much overclock - If anything is overclocked reset it to default speeds.

Bad power supply - no easy way to troubleshoot, just replace, especially if it too is from 2010.

RAM - I know you said you checked it but but how long and how? Download and run the latest Memtest86+ from a USB stick, the one included with MXLinux is old. I would let it run for at least 8 hours or more. Flakey RAM can take a long time to reveal. ALSO - 6GB is an odd configuration, usually you want identical memory sticks. Some systems are sensitive to this. If it used to be stable on 2GB RAM then remove the extra and try that.

A Bad SSD or SATA cable - especially if there are substantial system slowdowns and freezes that recover after a few seconds to minutes. Check the SMART data with GSmartControl. Post here if you want some help interpreting it.

Just an observation, but your CPU is pretty warm for what appears to be an idle system. When was the last time the system was opened and all the dust blown out, especially from fans and heatsinks?

More very good suggestions. The SATA one made me think of another: besides reseating the RAM modules, unplug then reconnect every single cable on the motherboard one at a time. Also do the cables connected to your disk drive. While you're in there blow out any dust that's there.

Another thing it could be on a computer that old is that the thermal paste between the CPU and its heat sink/fan has dried up and needs replacing. The lock-ups could be due to thermal shutdowns of the CPU due to insufficient cooling.

Anyway, you now have a big list of things to look into and things to try. That should keep you out of trouble for a couple of days or so.

[lostduckling]

Quick and Minor update:

Again, thank you so much for the responses. They really help me think and see new paths to learn. You turned an annoying problem into something exciting to figure out.

I'm unfortunately on a tight schedule this week, so I only have 1 hour of testing / learning free unless it is something very passive like turning it on and leaving it like that. I do plan to answer every single one of you specifically later, even if not to just add data to whoever needs in the future.

With that said I did narrow down it to hardware. I booted a flashdrive that runs windows 8.1 on it. It froze all the three times. Two near past login, once after a really long time after I decided to push 3 videos open on 3 different chrome windows (more on that later).

I'm yet to install whocrashed on the computer to see if I get more data out of the error, BUT on the time that it ran for longer before freezing the screen, I got a "Whea_uncorrectable_error" error, which is a generic hardware error, as if freezing on windows too wasn't enough.

As for now I ruled out both DDR3 memory sticks and both of the dual channel sockets (tested them all individually on linux).

I'll reset the bios between today and tomorrow, and see if I can find an update too.

I'm yet to open the CPU and add more paste to the heatsink.

Weird trivia that might point to integrated graphical failure:

Chrome (so, different resource usage on firefox) on windows ran for way longer, it only froze after I decided to push 3 windows running youtube for a long time. CPU and memory were doing fine, it is likely the chipset and processor dealing with it. Regardless it was longer than firefox.

If anyone is researching their own errors and seeing this: I'm not saying firefox is to blame, I had problems that didn't even use firefox. But like someone said, it is more resource intensive on older CPUs lately, no idea on chrome.

Before I go old GPU-hunting, any suggestions on how to better narrow down?

[j2mcgreg]

lostduckling wrote:

Before I go old GPU-hunting, any suggestions on how to better narrow down?

I think that you should open the case and observe the cooling fans when the machine is running Windows and when its running Linux. You want to determine if there is a noticeable difference in the fan speeds between the two operating systems. If the fans are running slower in Linux, you may be able to adjust the speed upwards by using the Smartfan utility in your bios.

The manual for your motherboard can be found here:

https://manualzz.com/doc/54555046/ecs-g ... ujcnBszQi9

[Mauser]

lostduckling, it has been my experience that when the computer freezes and or kernel panic it is due to failing hardware. The only other thing it can be is if you changed something in the BIOS. Updating the BIOS won't correct anything if it was working correctly earlier. I recommend you run Psensor to monitor the temperature of your hardware to see if anything is overheating. Overheating can cause random freezes and kernel panics. Sometimes opening the computer case and give it a good cleaning and if needed clean and reapply thermal paste on the CPU can resolve the issues you describe. Looks like you tried another drive so it is probably not a drive issue. Try running GsmartControl to verify that the drive is good. As mentioned earlier, re-seating RAM modules is a good idea. If the freezes and or kernel panic is failing hardware, there only way to figure out which hardware is causing the problem is trail and error by replacing components because no log file will tell you which hardware is causing the issue because when your computer freezes and or kernel panics it will stop writing logs. Eventually your computer won't boot up and you will know then which hardware it was causing you all the trouble like what happened to me once.

[lostduckling]

Guys, I know this is a late reply, but last few weeks have been quite busy to troubleshoot and open the PC in a way that attention would be properly gave to your thoughtful comments. Apologies.

I did take a few tests, cleaned it and put thermal paste, and it now seems to work as intended. It seemed to be a temperature issue.

PS: I'm still amazed by how many things i went on to go install on MX linux, and turns out it was already there. Incredible care and insight on this regard.

Thank you to EVERYONE that gave me a bit of their time, you saved me quite a bit of money and made me learn a lot.

I left the machine running for 2 hours on idle with barely a temperature change, no freezes. Then afterwards left it playing youtube, temperature went slightly up (about 8 C on the second core) but no problems whatsoever.

Thank you.

[lostduckling]

Sorry for double posting:

It wasn't temperature. Kinda.

The temperatures are very steady, and the computer can run for hours and hours while using cpu and memory, but randomly after a few hours now, it freezes.

Temperature certainly helped the problem, and not it is usable, but I have no idea what might have been.

I'll keep doing your tests and a few others a few friends suggested, but felt like you should know. Does that ring any bell?

[m_pav]

One thing not covered yet is your power supply. Bad or dirty power, whetrher too much or too little will cause random lockups that progressively get worse and can/will cause component failure, right up to and including the CPU.

Take a peek through the slots if you can and look for any bulging capacitors. If you have even one bulging capacitor, either have it fixed (replace all caps of the same type/value where possible) or switch it out for a good power supply. When a 1000uf Capacitor starts to bulge, it's measured capacitance value will be less then 1/4 of its original value, potentially down to 10%, so it will already be next to totally useless in its circuit causing unstable and dirty power.

Same goes for your motherboard. I've had to replace up to 15 caps on a motherboard in the past for a customers mission critical machine and in doing so, extended it's life by many years, but the key thing is, early intervention.

I encourage you to search for bulging capacitors to see just how rampant this issue is, especially with hardware in the age and date range of yours.

[spandey]

I have a Old PC like yours and used to have similar problem. Then I switched to X64 bit ISO rather than 32 bit ones and freeze problems are gone. System runs much smoother with 10 Firefox tabs open.

[xiloba]

When I used to use Slackware, I had to put append=" vt.default_utf8=0 intel_idle.max_cstate=1 acpi_osi=acpi_backlight=native" in lilo configuretions to avoid random freezes. My processor was n2930.