Revision of submit.slurm

1 file changed, 3 insertions, 1 deletion

README.md

			@@ -21,4 +21,6 @@ Run Jobs on NERSC
21	21
22	22		3.4. `<FULL_PATH_TO_RMCPROFILE_PACKAGE>` is the full path to the RMCProfile package directory.
23	23
24		-	3.5. `<RMC_RUNNING_DIR>` is the RMC running directory and `<YOUR_RMC_STEM_NAME>` is the stem name of the RMC setup
	24	+	3.5. `<RMC_RUNNING_DIR>` is the RMC running directory and `<YOUR_RMC_STEM_NAME>` is the stem name of the RMC setup.
	25	+
	26	+	4. Submit the job by running `sbatch submit.slurm` from where the `submit.slurm` script is located, on the terminal.

1 file changed, 1 insertion, 1 deletion

README.md

			@@ -9,7 +9,7 @@ Run Jobs on NERSC
9	9		tar xzvf RMCProfile_V6.7.9_Linux_64.tgz
10	10		```
11	11
12		-	2. Prepare the job submission script -- see the example above -- and put the script to somewhere in your home directory. I would suggest put the script in the same directory where the RMC run will be happening so that we can easily keep track of which one is for which one.
	12	+	2. Prepare the job submission script -- see the example below -- and put the script to somewhere in your home directory. I would suggest put the script in the same directory where the RMC run will be happening so that we can easily keep track of which one is for which one.
13	13
14	14		3. Several notes about the script,
15	15

2 files changed, 46 insertions

README.md(file created)

		@@ -0,0 +1,24 @@
1	+	Run Jobs on NERSC
2	+	===
3	+
4	+	1. First, download the RMCProfile package on Linux platform [here](https://yr.iris-home.net/rmclinux) and untar it somewhere in our home directory. Once logged onto NERSC from terminal, we can do this,
5	+
6	+	```
7	+	cd SOMEWHERE
8	+	wget https://yr.iris-home.net/rmclinux -O RMCProfile_V6.7.9_Linux_64.tgz
9	+	tar xzvf RMCProfile_V6.7.9_Linux_64.tgz
10	+	```
11	+
12	+	2. Prepare the job submission script -- see the example above -- and put the script to somewhere in your home directory. I would suggest put the script in the same directory where the RMC run will be happening so that we can easily keep track of which one is for which one.
13	+
14	+	3. Several notes about the script,
15	+
16	+	3.1. `<YOUR_JOB_NAME>` is a name that we can give to the running job, which can be any meaningful name to us.
17	+
18	+	3.2. `<YOUR_OUTPUT_FILE>` is the name for the output file of the running job, meaning that all the terminal output during the job running will go into this file.
19	+
20	+	3.3. `<YOUR_EMAIL>` is our email address -- status (e.g., started, or stopped, or failure, etc.) of the job will be sent to this email.
21	+
22	+	3.4. `<FULL_PATH_TO_RMCPROFILE_PACKAGE>` is the full path to the RMCProfile package directory.
23	+
24	+	3.5. `<RMC_RUNNING_DIR>` is the RMC running directory and `<YOUR_RMC_STEM_NAME>` is the stem name of the RMC setup

submit.slurm(file created)

		@@ -0,0 +1,22 @@
1	+	#!/bin/bash
2	+	#SBATCH --job-name=<YOUR_JOB_NAME>
3	+	#SBATCH --output=<YOUR_OUTPUT_FILE>
4	+	#SBATCH -t 24:00:00
5	+	#SBATCH -N 1
6	+	#SBATCH -q regular
7	+	#SBATCH -C cpu
8	+	#SBATCH --mail-user=<YOUR_EMAIL>
9	+	#SBATCH --mail-type=ALL
10	+
11	+	#OpenMP settings:
12	+	export OMP_NUM_THREADS=8
13	+	export OMP_PLACES=threads
14	+	export OMP_PROC_BIND=spread
15	+
16	+	RMCProfile_PATH=<FULL_PATH_TO_RMCPROFILE_PACKAGE>
17	+	export PGPLOT_DIR=$RMCProfile_PATH/exe/libs
18	+	export LD_LIBRARY_PATH=$RMCProfile_PATH/exe/libs
19	+	export LIBRARY_PATH=$RMCProfile_PATH/exe/libs
20	+	export PATH=$PATH:$RMCProfile_PATH/exe
21	+	cd <RMC_RUNNING_DIR>
22	+	$RMCProfile_PATH/exe/rmcprofile <YOUR_RMC_STEM_NAME>

更新更早