setup_instance_remote

#!/usr/bin/env bash
# This bash code will be ran as EC2 instance is initialized.

# OPTIONS

# R source url
r_url="https://cran.r-project.org/src/base/R-3/R-3.4.1.tar.gz"
# Rstudio Server source url
rstudio_server_url="https://download2.rstudio.org/rstudio-server-rhel-1.0.136-x86_64.rpm"

# list of all packages to be installed
packages="c('doMC', 'caret', 'data.table', 'plyr', 'dplyr', 'reshape', 'ggplot2')"
# CRAN repository
repos="https://cran.rstudio.com/"
# password needed to log in to rstudio
rstudiopassword="6kYMumDQ97fvRdbhp2eyAncMifDgWLwxZR"

# Amazon SSM URL (needed to run commands through AWS-CLI)
ssm_url="https://s3.amazonaws.com/ec2-downloads-windows/SSMAgent/latest/linux_amd64/amazon-ssm-agent.rpm"


################################################################################
# Since user-data scripts are run as root we do not use sudo
yum update -y

# Install SSM
cd /tmp
yum install -y $ssm_url

# yum R is outdated, instead installing R from source
cd /tmp
yum install -y gcc-c++ gcc-gfortran readline-devel libxml2-devel libcurl-devel bzip2-devel pcre-devel
wget $r_url
tar xfz "$(basename "$r_url")"
cd $(basename "$r_url" .tar.gz)
./configure --with-x=no --enable-R-shlib=yes --enable-java=no
make
make install

# Installing R studio server from Source
cd /tmp
wget $rstudio_server_url
yum install -y --nogpgcheck "$(basename "$rstudio_server_url")"

# Install all required R packages
R -q -e "install.packages($packages, repos='$repos')"

# Rstudio login over http (port 8787)
useradd rstudio
echo "rstudio:$rstudiopassword" | chpasswd

# Shutdown the instance when installation is finished
shutdown -h now