upload_video_list.sbatch

#!/usr/bin/env bash
#SBATCH --job-name=video2tator_OTZ
#SBATCH --ntasks=1
#SBATCH --cpus-per-task=32
#SBATCH --mem-per-cpu=2GB
#SBATCH --time=6:00:00
#SBATCH --partition=compute
#SBATCH --mail-type=ALL
#SBATCH --mail-user=
#SBATCH --array=
#SBATCH --output=slogs/video2tator/%A.%a.%x.txt

# LOGGING JOB DETAILS #
echo "Job ID: $SLURM_JOB_ID, JobName: $SLURM_JOB_NAME"
hostname; pwd; date

# SETTING UP ENVIRONMENT #
module load ffmpeg
module load gcc/8.5.0
source venv/bin/activate
umask 000  # newly created files have all permissions ok. 
echo "Environment... loaded"

set -eux  # exit on error, including unset vars

# Get target dir to from list and slurm-array-index to convert to video
VIDEOS_LIST="$1"
VIDEO_FILE=$(sed "${SLURM_ARRAY_TASK_ID}q;d" $VIDEOS_LIST)
VIDEO_NAME=$(basename "${VIDEO_FILE%.*}")

## TRANSCODING AND UPLOADING TO TATOR ##

echo "TATOR-PY..."
TOKEN=$(cat tator_token.txt)  # hpc-user access token
HOST=https://tator.whoi.edu
PROJ_ID=1   # ISIIS
MEDIA_ID=1  # shadowgraph video
FOLDER="$(basename $VIDEO_FILE | cut -f2 -d'_')"

time python3 -m tator.transcode "$VIDEO_FILE" \
    --name "$VIDEO_NAME" \
    --host "$HOST" \
    --token "$TOKEN" \
    --project $PROJ_ID \
    --type $MEDIA_ID \
    --section "$FOLDER" \
    --work_dir "tator_transcode_workspace" \
    --cleanup

echo 
TZ=UTC0 printf '%(%H:%M:%S)T\n' $SECONDS