icactus2003
Newbie level 1
lvheap memory allocation failure
I have a user running running Calibre xRC (parasitic extraction tool) on AMD Opteron machines (dual core).
After running for more than 10 hours, the process was killed by itself.
I noticed the server was "hang" before the process got killed. Not be able to ssh or login physically on the machine.
Under normal condition, the process should be completed within 2 hours.
Below is the message appeared on terminal:-
INFO: Starting to process net u_sub_top_apa_1/u_sub_top_apa_2/u_core_apa/u_to
p_arm/USB/sw_wire_2358_BF_u_sub_top_apa_1\_u_sub_top_apa_2\_u_core_apa\_
u_top_arm\_USB_U37192_Y LVHEAP = 760/761/763 MALLOC = 57124/57124/57124
INFO: Starting to process net u_sub_top_apa_1/u_sub_top_apa_2/u_core_apa/u_to
p_arm/USB/sw_wire_7601_BF_u_sub_top_apa_1\_u_sub_top_apa_2\_u_core_apa\_
u_top_arm\_USB_U31725_Y LVHEAP = 760/761/763 MALLOC 57127/57127/57127
Killed
25666.238u 876.815s 11:27:45.91 64.3% 0+0k 0+0io 225168pf+0w
I found the following information in the manual;
Troubleshooting Virtual Memory Allocation Failures
The most common memory allocation error you might encounter when using Calibre tools is LVHEAP ALLOCATION FAILURE.
In most cases, the Calibre tool issues this error because the platform's swap is insufficient. Generally, you should set the platform's stacksize no larger than 8192 kilobytes. You can verify this by using the limit shell command on both HP‑UX and Solaris platforms. You can temporarily modify the stacksize in a C shell by typing the following in the shell: limit stacksize 8192
But when user did the above method and resend the job the similar problem occured.
Please help....really no idea what has goes wrong here....
I have a user running running Calibre xRC (parasitic extraction tool) on AMD Opteron machines (dual core).
After running for more than 10 hours, the process was killed by itself.
I noticed the server was "hang" before the process got killed. Not be able to ssh or login physically on the machine.
Under normal condition, the process should be completed within 2 hours.
Below is the message appeared on terminal:-
INFO: Starting to process net u_sub_top_apa_1/u_sub_top_apa_2/u_core_apa/u_to
p_arm/USB/sw_wire_2358_BF_u_sub_top_apa_1\_u_sub_top_apa_2\_u_core_apa\_
u_top_arm\_USB_U37192_Y LVHEAP = 760/761/763 MALLOC = 57124/57124/57124
INFO: Starting to process net u_sub_top_apa_1/u_sub_top_apa_2/u_core_apa/u_to
p_arm/USB/sw_wire_7601_BF_u_sub_top_apa_1\_u_sub_top_apa_2\_u_core_apa\_
u_top_arm\_USB_U31725_Y LVHEAP = 760/761/763 MALLOC 57127/57127/57127
Killed
25666.238u 876.815s 11:27:45.91 64.3% 0+0k 0+0io 225168pf+0w
I found the following information in the manual;
Troubleshooting Virtual Memory Allocation Failures
The most common memory allocation error you might encounter when using Calibre tools is LVHEAP ALLOCATION FAILURE.
In most cases, the Calibre tool issues this error because the platform's swap is insufficient. Generally, you should set the platform's stacksize no larger than 8192 kilobytes. You can verify this by using the limit shell command on both HP‑UX and Solaris platforms. You can temporarily modify the stacksize in a C shell by typing the following in the shell: limit stacksize 8192
But when user did the above method and resend the job the similar problem occured.
Please help....really no idea what has goes wrong here....