Serious memory problem! Do I have a crapped one?

orangepiuser

I have running debian 8. It seems that I have a memory problem or something else.

When I do:

1. stress -v -c 2 -m 2 --vm-bytes 32M

Copy code

the system hangs after few moments like this:

1. stress: dbug: [577] touching bytes in strides of 4096 bytes ...
   
2. stress: dbug: [575] freed 33554432 bytes
   
3. stress: dbug: [575] allocating 33554432 bytes ...
   
4. stress: dbug: [575] touching bytes in strides of 4096 bytes ...
   
5. stress: dbug: [577] freed 33554432 bytes
   
6. stress: dbug: [577] allocating 33554432 bytes ...
   
7. stress: dbug: [577] touching bytes in
   

Copy code

Then I need to restart it by unplugging the power.

Can somebody test this to confirm or if I have a crappy board?

mrroland

do you have swap enabled? could be a full memory...

orangepiuser

I have set up a swap space, but still the same problem. Now I am not sure if its a memory problem or a cpu-problem with some kind of "overload".
Can somebody else try to stress his system and check if its a system problem or just mine?

orangepiuser

I have same stucking when I reduce the amount of memory thats tested. Even with vm-bytes=1 and from time to time even during startup. I tried three different SD-cards in different sizes.

It stops while doing apt-get update, too:

1. Get:3 http://ftp.hr.debian.org jessie-backports/main Sources/DiffIndex [5,059 B]
   
2. Get:4 http://ftp.hr.debian.org jessie-backports/contrib Sources/DiffIndex [367 B]
   
3. Get:5 http://ftp.hr.debian.org jessie-backports/non-free Sources/DiffIndex [781 B]
   
4. Get:6 http://ftp.hr.debian.org jessie-backports/main armhf Packages/DiffIndex [5,749 B]
   
5. Get:7 http://ftp.hr.debian.org jessie-backports/non-free armhf Packages/DiffIndex [229 B]
   
6. Get:8 http://ftp.hr.debian.org jessie-backports/main Translation-en/DiffIndex [1,609 B]
   
7. Get:9 http://ftp.hr.debian.org jessie-backports/non-free Translation-en/DiffIndex [367 B]
   
8. Get:10 http://ftp.hr.debian.org jessie/main Sources [7,058 kB]
   
9. Get:11 http://ftp.hr.debian.org jessie/contrib Sources [50.7 kB]
   
10. Get:12 http://ftp.hr.debian.org jessie/non-free Sources [99.4 kB]
    
11. Get:13 http://ftp.hr.debian.org jessie/main Translation-en [4,585 kB]
    
12. Get:14 http://ftp.hr.debian.org jessie/non-free Translation-en [72.2 kB]
    
13. Hit http://ftp.hr.debian.org jessie-updates/main Translation-en
    
14. Hit http://ftp.hr.debian.org jessie-backports/contrib armhf Packages
    
15. 100% [13 Translation-en bzip2 0 B] [10 Sources
    
16. 
    

Copy code

Now I downloaded 1.2GB with no problem using wget.

Copying 4GB via scp from a local server hangs the system after about half a minute.

WHAT is my problem, crazy, isn't it? Is there any command where I can put a "handbrake" on the system, to disable some cores?

mrroland

what type of board do you have? Orangepi - pc?

mrroland

Which script.bin and uImage are you using?

orangepiuser

I tried different script.bin and Uimages.

Now I found out something: I changed the power source to a bigger one. Now I can do the update via apt-get, but I still cannot do network-copies with scp. wget works. It seems that the machine needs a lot of power to run and when it does not get enough, it just hangs.
So I try to get an even stronger power supply and do more testings.

jacer

5V 2A should be OK.

makama80

I have more or less the same problem with my PCDUINO3 nano (Alwinner A20 based board). It hangs completely random and out of the blue. I've tried heavier power supplies, soldering capacitors over the 3V and 5V pins, extra large cooling block and fan, slowing down the SOC speed(s): nothing helps. Also no info on the screen appears what explains or helps finding the error: the device hangs completely.

On the other hand: for around $15 you have a new orange pi pc. I guess your board is an example of bad luck.... Maybe try to discuss the possibility of a refund / exchange at Xunlong.....

whitewind

Try to use memtester to isolate problem. It should be in distro repository.