I declare, I have seen the light although I was mired in doubt to start with.
What held me back from accepting the cloud into my life was fear - the terror of losing control.
It started when I joined an IT department on the bottom rung, as the support guy who installed a PC on your desk. From there, my role evolved into server applications, then networks and firewalls. In those days, experience was a brutal teacher, so you always needed a backup. We backed up everything before an upgrade because there was always a precedent for things going wrong.
People were fearful about moving forward. We even used that mindset to set user expectations by saying “If this goes wrong, your system could be down for the next three days”.
The problem was that servers had grown organically over many years, with patches applied incrementally to both operating systems and applications and, as staff changed, system changes were often poorly documented, if at all.
Inflexibility was the result. You could not move backwards, while moving forward - whether it was an application or infrastructure upgrade - was a lengthy and expensive process. The fear of something going wrong meant everyone had to be there and, because major changes typically occurred overnight at periods of low demand, everyone was expensive.
When the public cloud first appeared, I admit I personally found it hard to move away from that ‘backup first’ mindset. Back then, I was always on edge about what we’d do if something went wrong.
But after I joined UltraServe, where we build systems based on the concept of Immutable Infrastructure, I realised there really is a life beyond backups.
Phrases like Immutable Infrastructure and Blue-Green Deployment are no longer overhyped promises, but are daily parts of my life on this righteous path.
Being the world’s only hybris Platform as a Service (PaaS) provider, we support our clients and partners with a product suite that enables them to get to market faster and then provide ongoing support to guarantee the outcomes that suit their individual needs. Vendors such as hybris, Amazon Web Services (AWS) and New Relic give us the tools required to build this platform and provide the best service and offering in the market today.
If you open the hood of our PaaS offering, SmartStack, what underpins this is Immutable Infrastructure, which is the precise opposite of the ‘house of cards’ architecture that instilled such fear when I started out.
While there’s no standard definition of Immutable Infrastructure, it describes a model where an entire software environment never changes once it is deployed: If you want to make changes, you replace its entirety with a new instance.
The process of Blue-Green Deployment is a natural consequence of Immutable Infrastructure. Under this model, “Blue” represents the currently deployed version of your software infrastructure while “Green” is the version intended for deployment.
We always move from Blue to Green - even where Green is a previous software state.
Whereas in the old days, we were tied to server infrastructure, today it is done in software. This allows the continuous evolution of everything from the operating system and applications to supporting programs such as Java.
If a deployment does not work out or unexpected consequences occur, we simply destroy the new environment and deploy the previous ‘good’ environment.
The main advantage for customers is that their e-commerce infrastructure now evolves much quicker with many fewer errors.
As a result, our SmartStack customers are seeing the smoothest greenfield deployments of code ever, achieved with a minimal number of people.
This means UltraServe can deliver a service that is significantly cheaper than was possible before because customers no longer need Emergency Services around in case there’s a train wreck.
It’s also much quicker. Whereas in the old days, the process of putting new functionality into an application could take weeks, months or even years in an exorbitantly expensive project, now, customers can put new functionality in place in days or even hours - again and again!
While we’re still in the early days, UltraServe has customers at the bleeding edge: They say this new world is not frightening. The old world was much scarier.
Some companies still take the old school approach with big teams and large budgets, but in reality, anyone who still provisions their own infrastructure needs their heads read.
Today, the cloud is a legitimate platform. As Fortune 100 companies publicise their successes with Immutable Infrastructure and Blue-Green Deployment, that news will filter out into the market and increase the adoption rate, which is just fine with us.
Because, like many converts before me, I am a now true believer.
- The three elephants in the room with the NBN At this month’s Commsday Summit in Melbourne, Internode managing director Simon Hackett warned of “three elephants” that threaten to shake up effectiveness of the National Broadban...
- 2010 is the year of IPTV for Australia At last, 2010 is the year of Internet Protocol Television (IPTV) for Australia – due to the convergence of on-demand video content and fast network access. After years of threaten...
- Broadband: A decade of ADSL in Australia Simon Hackett argues that infrastructure-based competition is an esssential condition for Australians to receive affordable access to high-speed broadband nationally. Simon, the ma...
- Bad habits are worse than bad guys in IT security IT security is at greater risk from the bad habits of system administrators than from bad guys actually hacking into infrastructure warns Centrify Asia-Pacific Regional Director Ma...