← Apache Spark learning pathApache Spark · Module 3 of 10
RDDs, transformations and actions
Learn Spark's core abstraction and lazy evaluation.
📖 Key ideas
- RDD: Resilient Distributed Dataset, an immutable distributed collection.
- Transformations (map, filter) build a plan lazily.
- Actions (count, collect, save) trigger execution.
- Lineage lets Spark recompute lost partitions.
- DataFrames are preferred today for most work.
💡 Real-life example: Lazy evaluation is writing a full shopping list before going to the market once.
RDDResilient Distributed Dataset
TransformationLazy operation creating a new dataset
ActionOperation that triggers computation
LineageRecord of how data was derived