An overview of AWS networking
For those of you who are familiar with general networking principles, AWS will be a substantially different beast from what you are used to. Within an enterprise you will be familiar with IP subnets and routing between those subnets. Perhaps each floor of your office has a different subnet, or perhaps each switch in your data center has its own Virtual Lan (VLAN), but whatever your current setup is, it will most likely look something like the following:
Each server is clearly on its own subnet, segregated into VLANs and routed by dedicated routers.
However, within AWS, networking looks substantially different.
Firstly, IP addresses are allocated automatically and change every time a server is restarted. Servers on AWS exist in a pool of servers, which can be located on any subnet within that availability region. Routing is managed by Amazon, and it is possible that there may be 1+n routers between all of the servers within your application:
Note
This is different...