>file /sbin/init
/sbin/init: ELF 64-bit LSB shared object, x86-64,...
>lsb_release -a
Description: Ubuntu 13.10
Release: 13.10
Codename: saucy
> cat /etc/*-release
DISTRIB_ID=Ubuntu
DISTRIB_RELEASE=13.10
DISTRIB_CODENAME=saucy
DISTRIB_DESCRIPTION="Ubuntu 13.10"
NAME="Ubuntu"
VERSION="13.10, Saucy Salamander"
...
>/sbin/ifconfig - shows Bcast address, e.g. 192.168.1.255
>ping -b 192.168.1.255 - shows other machines on local subnet
cd ~/.config/Slack
rm -Rf SS
pandoc in.md -o out.pdf
pandoc in.md -o out.html
pandoc in.md -o out.html -s --metadata pagetitle="Out"
# Updated for pandoc 2.4 - making slides
DOCNAME=SparkDataEngineering
pandoc -t beamer --pdf-engine=xelatex ${DOCNAME}.md -o ${DOCNAME}.pdf -V theme:metropolis -V colortheme:default
# Slides
---
header-includes:
- \usepackage{fvextra}
- \DefineVerbatimEnvironment{Highlighting}{Verbatim}{breaklines,commandchars=\\\{\}}
- \usepackage{fontspec}
- \usepackage{setspace}
title: Data Engineering with Apache Spark
author: Markus Dale, medale@...
date: May 2019
---
# Apache Spark: Data engineering for larger dataset (Vertical Scaling)
{height=80%}
# Cluster Manager - Manage cores, memory, special capabilities

# Data Science Mission
\Large
* ID malicious GitHub Pull Requests?
# Anatomy of a Spark Application

\tiny Source: Apache Spark website
# Hello, Spark World!
\scriptsize
scala (code block with three backticks)
import org.apache.spark.sql.SparkSession
...
xdg-open steps.html
xdg-open http://cnn.com