Seven Concurrency Models in Seven Weeks

Day 3: The Speed Layer

As we saw yesterday, the batch layer of the Lambda Architecture solves all the problems we identified with traditional data systems, but it does so at the expense of latency. The speed layer exists to solve that problem. The following figure shows how the batch and speed layers work together:

Figure 22. The Lambda Architecture

As new data arrives, we both append it to the raw data that the batch layer works on and send it to the speed layer. The speed layer generates real-time views, which are combined with batch views to create fully up-to-date answers to queries.

Real-time views contain only information derived from ...