Disposable Infrastructure

The more we build up our tool set for scaling out the applications we build for both our clients and for ourselves, the less these apps are dependent on any single piece of infrastructure. The only part of our application infrastructure we really manage now are the application servers because everything else is provided as a service by Amazon or other providers including the database.

For those application servers, we have tools to manage them so easily now that we don't really worry about scaling out our apps, nor do we worry if they crash. Need more capacity? Launch more servers. Server crashes?  Launch a new one. Launch two if it makes you feel warm and fuzzy. Launched too many? Shut em down. We even launch new servers just to update a piece of software on them. It is literally easier to launch a new server that configures itself entirely including checking out the application code and adding itself to the load balancer, than it is to ssh in and update something on the server even with apt-get. In fact, I find having to ssh into a server an unnecessary pain now. Regularly killing servers and replacing them with fresh new ones is what I like to call Disposable Infrastructure.

Now that we at Appoxy have the experience, knowledge and tools to make managing and scaling applications so easy (and not scary), we firmly believe that this is the only way things should be done. It wasn't long ago that this type of thing wasn't even in the realm of possibility. Amazon launched EC2 in August of 2006 and this was really the game changing moment. Before then you either bought your own hardware or you leased servers by the month. EC2 brought that down to leasing by the hour which is what makes the disposable infrastructure a sane (ie: not-so-crazy) reality.

If you're interested in seeing how we do this, please get a hold of us, we'd be happy to show you.