Get a database dump via psql and push to S3

psql-s3-backup.sh

#!/usr/bin/env bash

if [ "$1" == "" ]; then
    echo "Bucket name required"
    exit 1
fi

DATE=$(date +%Y%m%d_%H%M)
DB_NAME=$PGDATABASE
BUCKET=$1
PSQL_HOST=$PGHOST
PSQL_USER=$PGUSER

if [ "$PSQL_HOST" == "" ]; then
    PSQL_HOST=localhost
fi

if [ "$PSQL_USER" == "" ]; then
    PSQL_USER=postgres
fi

# Execute a database lookup
(psql -U "$PSQL_USER" -h "$PSQL_HOST" -lqt | cut -d \| -f 1 | grep -qw "$DB_NAME")
EXISTS=$?
if [[ "$EXISTS" != 0 ]]
then
    echo "Database does not exist"
    exit 1
fi

echo "Database exists. Preparing dump"

OUTPUT_FILE="/tmp/$DB_NAME-dump_$DATE.sql.bz2"

pg_dump -U "$PSQL_USER" -h "$PSQL_HOST" "$DB_NAME" > "/tmp/$DB_NAME-dump_$DATE.sql"

bzip2 -c "/tmp/$DB_NAME-dump_$DATE.sql" > "$OUTPUT_FILE"

aws s3 cp "$OUTPUT_FILE" "s3://$BUCKET/dumps/"

# Remove temporary files
rm "$OUTPUT_FILE"
rm "/tmp/$DB_NAME-dump_$DATE.sql"