DevOps Talks Conference, 2017

A light dew settles on the leaves of the venerable elm trees which track Melbourne’s, St Kilda road. The bulk of the noble Yarra River moves majestically past the Melbourne Convention and Exhibition Centre, as it does every morning unaware about what is about to take place within. Light rail service number 96, which I had boarded at the iconic Luna Park, will carry me on my sovereign journey to the DevOps Talks Conference, 2017.  

Yet still, those lingering words circle around my mind, like plastic bags caught in the wind, waiting to be sucked into a stormwater drain. Self-doubt is setting in. Am I clearly delusional?

“You’re going to a DevOps conference? Aren’t you a developer?”

This is something I had been asked on more than one occasion in the lead up to this conference. Each time I’m questioned, I point out that the term DevOps is exactly six characters long, and that more or less fifty percent of those characters are “Dev”. I have at least half a right to be here.

Will Swift be the next king of server side development?

Swift throne

In June 2015, Apple announced at WWDC that they were open-sourcing the Swift language and its runtime libraries. On December 3rd that year they made good on their promise. In this post I’d like to talk about why this is significant, particularly for server-side developers.

Google Cloud Dataproc and the 17 minute train challenge


My work commute

My commute to and from work on the train is on average 17 minutes. It’s the usual uneventful affair, where the majority of people pass the time by surfing their mobile devices, catching a few Zs, or by reading a book. I’m one of those people who like to check in with family & friends on my phone, and see what they have been up to back home in Europe, while I’ve been snug as a bug in my bed.

Stay with me here folks.

But aside from getting up to speed with the latest events from back home, I also like to catch up on the latest tech news, and in particular what’s been happening in the rapidly evolving cloud area. And this week, one news item in my AppyGeek feed immediately jumped off the screen at me. Google have launched yet another game-changing product into their cloud platform big data suite.

It’s called Cloud Dataproc.

SSH through a Raspberry Pi Railway Signal

lightipadWe’ve all been there. You are in the supermarket with two bottles of diet cola in one arm and a packet of brown rice with quinoa in the other. Your site lead calls you with a request from a client who has locked themselves out of their account. Normally you would direct them to the administration interface but, because of the paradox inducing way in which they have bent the fabric of space and time, this will require some manual intervention.  You need to apply some subtle but distinct database changes. Simply delete a row or two from the QUANTUM_PARADOX table. Well, it’s actually a view… but that’s not important right now.

RHEL 5 – Xen & GFS

Last week I went to a very interesting presentation in Melbourne by Redhat and HP about the release of RHEL 5. This does seem to be quite a step forward for RedHat and Enterprise software – and it was quite noticable that the meeting had generated a great deal of interest (the room which could only really hold 50 people was jam packed with about 70). There were some good speakers, some of whom had been shipped over from america by Redhat who seem to have some genuine experiance with RHEL out in the field with some very big clients (washington area / stock exchange). There were two areas of technology which i believe could be very important

1) GFS – The Red Hat Global File System (GFS) is a filesystem like ext3 etc but lets many servers share a file system and seemed very powerful – the idea behind it is providing support for cluster storage

2) Xen – They have fully integrated virtualization into RHEL 5, providing kernels, gui support etc. Xen supports more than linux virtualisation including support for windows XP etc. More information here

Using RHEL5, Clustering, GFS and Xen large amounts of processing power can be made available across multiple cluster nodes on varying hardware. Providing for example, multiple test environments without needing separate dedicated hardware or zero downtime for clients – the “servers” can be migrated live between nodes with users experiancing only a slight lag on some network requests (this was bravely demo’d at the presentation and seemed to work well).