https://bugzilla.novell.com/show_bug.cgi?id=655221
https://bugzilla.novell.com/show_bug.cgi?id=655221#c2
John Sørensen changed:
What |Removed |Added
----------------------------------------------------------------------------
CC| |john@hovedpuden.dk
--- Comment #2 from John Sørensen 2011-01-14 12:35:48 UTC ---
(In reply to comment #1)
I have a similar problem on two systems both running openSuSE 11.3.
One of the systems has 96GB of memory and is used for R jobs running in 5
parallel processes each having 10GB data.
On this system the problem occurred after adding two swap files of 10 and 20GB
both residing on the same ext4 root file system.
One way of reproducing this problem is:
1. create 2 swap files of 10GB each on an ext4 file system
Filename Type Size Used
Priority
/dev/sda1 partition 2096476 2096476 -1
/root/swap file 10485756 418692
-2
/root/swap1 file 10485756 418672
-3
2. run some file system intense jobs like tar -cf /dev/null /usr
3. run some memory intensive jobs starving the configured swap.
An example of memory starving jobs might be running R jobs in parallel mode or
this executing simple code in an endless loop for around 2 hours on a system
with 16GB of memory and two 10GB swap files in an ext4 file system:
#include
#include
#include
int main(int argc, char** argv) {
char* buffer;
buffer=malloc((long)10*1024*1024*1024);
fork();
memset(buffer, 0, (long)10*1024*1024*1024);
printf("Allocated %d MB\n",10*1024 );
return 0;
}
Depending on the size of your memory more instances of the program might be
needed.
--
Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are on the CC list for the bug.
