Chapter 5. Fault Tolerance and Reliability
Apex was built from the ground up for enabling data processing pipelines that are highly available and providing strong processing guarantees for accurate results. From its first release in 2013, Apex has supported exactly-once semantics based on distributed checkpointing and full fault-tolerance, including the application master. In this chapter, we will look at how these important properties are achieved, and what different components in the stack contribute to it:
- The need for the distributed data processing platform to be resilient
- Failure scenarios in Apex and how they are handled
- Consistent, distributed checkpointing and how it works
- Efficient, incremental, and large-scale state saving
- Why the accuracy of the processing is important, and what guarantees does Apex provide
- An example application for end-to-end exactly-once results
