Open source architecture
It all starts with an idea
When you sit down and start to think about it, we, as humans, are very good at learning new things and using our understanding to progress and do new things based on what we learnt. The founding principle is everything I know can be expressed thusly:
“What I know” = (“What I learnt” – “What I forgot about what I learnt”) + My experiences
So based on this every time I do something new I learn something, I then forget a proportion of that but have hopefully formed a subconscious opinion of it and have now gained an experience of that.
One of the ways I learn more is to read, use the internet to find solutions and adapt them to my current situation, sometimes that results in an experience or something I’ll actually retain and will re-use at another time.
Now when I’m often learning I am also working so I apply the “what I know” to the situation I am in with a dash of learning and experience and with some healthy collaboration end up at a solution, now isn’t that the key? collaboration is when we start sharing our “what I know” and our experiences of it to come to a solution that often removes the pain points that each member experienced in the past but potentially introduces more.
So that leads me on to the idea, what if we open source our architecture? Does that provide value? I think so, we have spent months years maybe on our architectures getting to a point that we as individuals are happy with them for various solutions, how cool would it have been if you could go to google and type “Open source architecture for hosting java based cloud solutions” and have it come back with something other than adverts for cloud solutions, maybe diagrams of the solution, maybe configurations, guidelines, bullet points directions of use etc etc. Wouldn’t that be like collaborating with others on your architecture, getting more experience and more knowledge to discuss the idea and suggest improvements. All of a sudden your one or two person IT team is now 30 or 40 people discussing the architecture, if just one of those people comes back with something that isn’t crap and is beneficial doesn’t that make it worth while?
Is there value in it?
I keep asking myself this. If every time I wanted to do a new “solution” there were “boxed” architectures for hosting applications in Amazon or setting up a data centre infrastructure I could pick it off the shelf, read some foot notes or supporting documentation understand the limitations and contraints of the solution and then adapt it to meet my specific needs.
I may even discover something that specifically could be contributed back to help out. For example, maybe Amazon EBS volumes become increasingly unstable so I fork the architecture and release an instance store version that then removes the dependancy on EBS or maybe a bug is found with ELB’s and access that causes a security hole, what ever it may be there a continuous feedback loop enabling everyone to benefit from the “what I learnt” and “My experience”
Okay, so in theory it seems okay, but, everyone has a unique situation so what ever is open sourced is potentially a waste? I don’t think so. Most of the principles behind what is done in any environment are the same, they are just tailored, I imagine most Corporate data centres have a number of distinct networks, probably a DMZ for web facing applications, maybe a more standard Application network for everything else that is not public facing or is internally accessible and lastly a secure area for DB servers or other important things that can never be accessed directly.
I think that’s pretty common, so the Open source architecture would depict this with maybe a dotted line to show A.N.Other network, maybe for Development, or for Testing but it’s a good starting point. I think the starting point coupled with the usage notes, diagrams design principles and documentation should be enough to get people up and running and with a decent architecture.
I’m sure there’s something in this, I know it will be beneficial, I worry that people will switch off and not take the architectures and adapt them to their specific needs, I hope in most cases that won’t be necessary, but there will always be fringe cases. Hopefully over the coming months this will turn into something more than words, but only time will tell.