ovro-lwa-solar-ops

Scripts and codes related to operations of OVRO-LWA for solar studies

Realtime Pipeline

Slurm managed run

By default, realtime pipeline runs on 10 nodes from calim cluster, every node requires 12 cpus and 32G Mem.

#!/bin/bash
#SBATCH --job-name=solar-test
#SBATCH --partition=general
#SBATCH --ntasks=10
#SBATCH --ntasks-per-node=1
#SBATCH --cpus-per-task=12
#SBATCH --mem=32G
#SBATCH --time=14:00:00
#SBATCH --output=/lustre/solarpipe/slurmlog/%j.out
#SBATCH --error=/lustre/solarpipe/slurmlog/%j.err
#SBATCH --mail-user=pz47@njit.edu

To start (restart). First check the existing running job with sinfo, then scancel if necessary. Then:

sbatch runSlurm_solarPipeline.sh slow # for slow pipeline
sbatch runSlurm_solarPipeline.sh fast # for fast pipeline

To run slow imaging for given period of time, do:

sbatch runSlurm_solarPipeline.sh testslowfixedtime \
 "2024-09-15T20:00:00"  "2024-09-15T21:00:00" 

Alignment to sunrise time

Use sunwait (https://github.com/risacher/sunwait) in crontab to submit slurm script.

Need to git clone download source file.

The crontab command for daily schedule.

0 10 * * * sunwait wait rise 37.2332N 118.2872W && sbatch /lustre/peijin/ovro-lwa-solar-ops/runSlurm_solarPipeline.sh slow