In the last chapter, we discussed the Fan-out Pattern, which we can implement using different strategies. At the end of that section, we reviewed an implementation of the Fan-out Pattern, which used AWS's Simple Queuing Service (SQS) as a destination for an event trigger. Queuing systems such as SQS provide a level of safety and security because they're intended to be a mostly durable persistent store where data lives until some process has the chance to pull it out, perform some work, and delete the item. If a downstream worker processes a crash entirely and processing stops for some time, queues merely back up, drastically reducing the risk of data loss. If a worker process runs into some unrecoverable problem in the middle of processing, queue items will typically be left on the queue to be retried by another processor in the future.
In this chapter, we will cover using queues as messaging systems to glue together multiple serverless...