Most business data is produced as a sequence of events, or an event stream: for example, web or mobile app interactions, devices, sensors, bank transactions, all continuously generate events. Even the current state of a database is the outcome of a sequence of events. Treating state as the result of a sequence of events forms the core of several event-driven patterns.
Event Sourcing is an architectural pattern in which the state of the application is determined by a sequence of events. As an example, imagine that each “event” is an incremental update to an entry in a database. In this case, the state of a particular entry is simply the accumulation of events pertaining to that entry. In the example below the stream contains the queue of all deposit and withdrawal events, and the database table persists the current account balances.
The events in the stream can be used to reconstruct the current account balances in the database, but not the other way around. Databases can be replicated with a technology called Change Data Capture (CDC), which collects the changes being applied to a source database, as soon as they occur by monitoring its change log, turns them into a stream of events, then applies those changes to a target database. Source code version control is another well known example of this, where the current state of a file is some base version, plus the accumulation of all changes that have been made to it.
What if you need to have the same set of data for different databases, for different types of use? With a stream, the same message can be processed by different consumers for different purposes. As shown below, the stream can act as a distribution point, where, following the polygot persistence pattern, events can be delivered to a variety of data stores, each using the most suited technology for a particular use case or materialized view.