Hi!
I experienced a similar problem. I tried burnmmx from the cpuburn package.Hi, So just goes to show how difficult it can be to prove that ram is at fault. Glen
"burnmmx P >> burnmmx.results &" starts one burnmmx process which uses 64 MB memory.
Start about as much processes as needed to use up all memory. If you want to test swap as well, use more.
Burnmmx stresses the cpu (if you have cooling issues) and checks if the data is still right every 512MB of calculation output, so you also test your ram.
For every process that terminates with corrupted data, you will find a "254" error code in burnmmx.results. If you see any of those, you have either a cpu or ram or cooling problem.
In one machine I had 1,25 G of SDRam-133. The board had an FSB of 100 MHz, but supported 133MHz ram speed. Memtest never showed any errors, but my 30 burnmmx processes all died within 30 minutes. Lowering the ram clock to 100MHz in BIOS manually solved the problem.
Hope this helps (others as well), Norman Schmidt.
glen schrieb:>
_______________________________________________ Ext3-users@redhat.com https://listman.redhat.com/mailman/listinfo/ext3-users
-- -- Norman Schmidt Institut für Physikal. u. Theoret. Chemie cand. chem. Friedrich-Alexander-Universitaet schmidt@naa.net Erlangen-Nuernberg _______________________________________________ Ext3-users@redhat.com https://listman.redhat.com/mailman/listinfo/ext3-users
