Home

Introduction

Based on:

• https://training.linuxfoundation.org/certification/lfcs

Pre-reqs

• Github account
• Reasonable package management on the host machine: on MacOS X, http://brew.sh
• Able to use vim (finish vimtutor) and shell/terminal (bash and host shell)

Environment setup (Virtual Machine)

We've selected Ubuntu 14.04 as our preferred Linux distribution.

We recommend vagrant to create VM:

• Install vagrant (https://www.vagrantup.com/docs/installation/)

On MacOS X:

$ brew cask install virtualbox
$ brew cask install vagrant
$ brew cask install vagrant-manager

• vagrant box add ubuntu/trusty64 (this installs the image for Ubuntu 14.04)

• in a new directory:

vagrant init ubuntu/trusty64
vagrant up 
vagrant ssh

Project

Overall goal:

• Setup 3-tier WebApp architecture and optimize:
• Webserver: (NGINX)
• Application server (NodeJS)
• database (PostgreSQL)

...also, add data analysis (Jupyter).

Meta: collect, analysis, report on data generated by the system itself (logs and monitoring information)

Overview of planned learnings

• optimize for China (packages from Chinese mirrors)
• configure timezone for China (make sure to use NTP)
• security
• automate setup and installation
• test different workloads
• benchmark and tune operating system and each tier
• depending on simulated load and hardware profile, select:
• sysctl settings tweaks
• service configuration and optimization
• filesystem / configuration
• VPN setup
• services monitoring
• load monitoring (memory, disk, cpu)
• log monitoring
• networking configuration
• firewall configuration

Exercises

Week 1: installing/configuring the webserver

• setup and install an instance of Ubuntu 14.04 LTS Server with Internet connectivity
• list all services running
• update packages source
• install webserver: nginx
• list all services running
• list files installed by nginx using the package manager
• make sure that the webserver is up, running and listening to port 80
• from your own host and using your browser, access http://localhost and make sure the above serves the nginx welcome page
• change the webserver configuration to serve your own index.html file from your home directory

Week 2: installing/configuring the database server and data source

• install the database server: postgresql
• make sure that the database server is up
• create a database
• using port forwarding, access this database using the DB GUI from your workstation
• install rsyslog-pgsql package
• http://shenavandeh.blogspot.com/2013/01/how-to-make-rsyslog-to-write-syslogs-in.html
• modify /etc/rsyslog.d/pgsql.conf to look as follows:

$ModLoad ompgsql
$WorkDirectory /var/tmp/rsyslog/work
$MainMsgQueueFileName mainq
*.* :ompgsql:localhost,rsyslogpgsql,rsyslogpgsql,<this is the generated password>

• create directory from the config file above: mkdir -p /var/tmp/rsyslog/work
• restart rsyslog service
• query database rsyslogpgsql and table systemevents for count of events
• create syslog events at each priority level
• query systemevents table for last ten messages

Week 3: installing the app server

• install vm nodejs
• create appserver user
• add code
• configure nginx to point to appserver

Week 4: separating the app into separate machine instances

• use centralized logging for all instances

Week 5: automation

• use ansible to automate the installation so far