Static screening error with mpirun
Posted: Tue Feb 20, 2024 2:01 am
Hello Yambo experts:
I hope you are all dong well today. My name is Xin Chang at UT austin. I am running Yambo on monolayer MoS2. I am following the tutorial on the Yambo wiki. I have already initialized the database without any error reported. Then, I used the command listed on the webcite, yambo -X s -F 01_3D_BSE_screening.in . Then, in the input file, I only changed these two settings
% BndsRnXs
42 | 65
% LongDrXs
1.000000 | 1.000000 | 0.000000 |
When I submitted the job, I used one of my scripts in the PBS job system with 1 node 24 cores. mpirun -np 24 yambo -F 01_3D_BSE_screening.in -J 3D_BSE
Later, in the folder, the Yambo package generates 18 outputs flies, with the name l-3D_BSE_screen_dipoles_em1s and l-3D_BSE_screen_dipoles_em1s_01 .... Then, I keep an eye on the outputs. I found there will be an error before step [04] dipoles in most outputs, which says "[ERROR] STOP signal received while in[04] Dipoles [ERROR] Writing File ./3D_BSE//ndb.dipoles; Variable NOT DEFINED; Permission denied". But, the interesting point is that the job is still running. I found there are two outputs (I have tried several times, each time the name of the outputs is differetn. ), let's say, l-3D_BSE_screen_dipoles_em1s_09 and l-3D_BSE_screen_dipoles_em1s_02 are still updating. They did not receive the errors. And they go to the step 04 and step 5 to calculate the dielectric constant. It takes a while (generally, 40 mins) to finish the calculation. And I can still run the next step calculation with the files in the folder, 3D_BSE.
I am not sure if you get my idea. It seems that the code only allows 2 processes to run the calculation. Please let me know what I should do to avoid that error. Best regards,
Xin
I hope you are all dong well today. My name is Xin Chang at UT austin. I am running Yambo on monolayer MoS2. I am following the tutorial on the Yambo wiki. I have already initialized the database without any error reported. Then, I used the command listed on the webcite, yambo -X s -F 01_3D_BSE_screening.in . Then, in the input file, I only changed these two settings
% BndsRnXs
42 | 65
% LongDrXs
1.000000 | 1.000000 | 0.000000 |
When I submitted the job, I used one of my scripts in the PBS job system with 1 node 24 cores. mpirun -np 24 yambo -F 01_3D_BSE_screening.in -J 3D_BSE
Later, in the folder, the Yambo package generates 18 outputs flies, with the name l-3D_BSE_screen_dipoles_em1s and l-3D_BSE_screen_dipoles_em1s_01 .... Then, I keep an eye on the outputs. I found there will be an error before step [04] dipoles in most outputs, which says "[ERROR] STOP signal received while in[04] Dipoles [ERROR] Writing File ./3D_BSE//ndb.dipoles; Variable NOT DEFINED; Permission denied". But, the interesting point is that the job is still running. I found there are two outputs (I have tried several times, each time the name of the outputs is differetn. ), let's say, l-3D_BSE_screen_dipoles_em1s_09 and l-3D_BSE_screen_dipoles_em1s_02 are still updating. They did not receive the errors. And they go to the step 04 and step 5 to calculate the dielectric constant. It takes a while (generally, 40 mins) to finish the calculation. And I can still run the next step calculation with the files in the folder, 3D_BSE.
I am not sure if you get my idea. It seems that the code only allows 2 processes to run the calculation. Please let me know what I should do to avoid that error. Best regards,
Xin