Yesterday, I gave my first talk at Infrastructure Coders on the Why, What and How of RatticDB.

Besides being slightly underprepared for it, it went really well and got some really positive feedback and geniune interest. Anyway, these are the slides from last nights talk:

On Monday I announced that I was leaving Luma Pictures, so this means you sir - Yes, you could have your very own linux systems administrator!

• 2.5 years experience
• studying bachelor of technology at RMIT University
• project lead on Rattic
• also batman.

Why me? Besides being a comedic genius, I’d like to think I’m a pretty smart lass with some pretty awesome war stories - data analytics, film and managed hosting. During this time I have got to design in-house tools with Django, manage over 4PB of storage and help developers see their dream come to life in AWS.

The biggest thing I’m selling is my passion for my career, I work as a systems administrator because I love problems, I love working things and making those things better - IT is forever changing, there is always a better way to do something.

But less about me, more about you - I want to love you, in the non-lawsuit kind of way. I want to work for a company that I love, people who are passionate, smart and constantly working on doing the best for themselves and the company.

If you think I may be a fit for your company, give me a hollar on twitter or by e-mail.

These days most companies are using configuration management of some sort, ansible, chef or in this case Puppet - which is great. Soon, snowflake systems will be a thing of the past and just a story that old greybeard Ben will tell new-comers.

At my work, I recently encountered a module which was unreliabe and unreadable. Worst thing? It wasn’t our only module that suffered this. After much swearing we decided that it was time to fix this and make our enviroment sustainable, here is a list of how to be less of douchebag:

• Choose a styleguide, stick to it and make sure everyone is on the same page.
• Documentation, I shouldn’t need to explain this.
• Code review.
• Dynamic scoping is bad and you should feel bad for using it.
• Read this

  With these at hand, we’re starting to make a positive step towards being able to manage our infrastructure without the pain and we are finding that with this new simplicity that we are finding new pain points and able to grow with ease.
  Agree, disagree? Tweet me @zemmiph0bia

I made the decision to leave WordPress, after trying multiple outlets I finally decided on Octopress, self-proclaimed ‘blogging framework for hackers’.

Why? I think the better question is “Why not?”

• It’s fast, wordpress is grossly over-featured for my needs.
• Losing the database means I removed my single point of failure.
• Simplicity laziness - All I need is vim and I can backup to GitHub.

I think it’s a given that if you aren’t comfortable with the concepts mentioned above, I’d probably give this a miss but if it excites you in the naughty way I’d certainly give it a gander.

I recently wrote a nagios check to monitor the status of our YellowFin 5 license due to it always running out during production hours. This check was written for version 5, I believe that they have made it easier for people to keep on this (and providing more than 90 day licenses)

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20

#!/bin/bash

# Settings
URL="http://urltomyyellowfin/info.jsp"
WARN=5
CRIT=1

PAGE=`wget $URL -qO - `
EXPIRYLINE=`echo "$PAGE" | tr "\n" " " | grep -o "<td>Expiration Date:</td>\s*<td>.*</td>\s*<td>\s*[0-9]* Days Left" | grep -o '[0-9]* Days Left'`
EXPIRY=`echo "$EXPIRYLINE" | sed 's/\([0-9]*\) Days Left/\1/'`

if [ "0$EXPIRY" -lt "$CRIT" ]; then
  echo "CRITICAL 1 Day Left"
  exit 2
elif [ "0$EXPIRY" -lt "$WARN" ]; then
  echo "WARNING 5 Days Left"
else
  echo "OK"
fi
echo $EXPIRYLINE