How is state managed between sessions? I know that in Azure, client-specific states are stored in SQL Azure. I’m wondering if this is done similarly in AWS?
Do the various instances of your application all access a DB somewhere where the state is stored? Is state management much different depending on which technologies you are using?
At a ‘homework’ level, Amazon Web Services is loosely comprised of two different sets of things:
When you upload a file to S3, it is stored across a number of machines in a number of different data centers, and Amazon is responsible for finding and returning the file when you request it (as well as making sure it doesn’t get erased by a machine failure.)
With something built on top of one of the infrastructure services, such as an application running on EC2, you are on your own as to how you store and synchronize state:
How do you store state? Traditionally a database (possibly Amazon RDS) with a memory cache (such as Elasticache – Amazon’s managed memcached-compatible cache). Amazon’s new DynamoDB service is a good fit for this use, as a fast, redundant, key-value store.