Caching of DataFrame between processing phases
Abierto
cloud
Conformance
feature
priority: undecided
Standardization
under discussion
- Lenguaje dominante
- Scala
- Estrellas
- 33
- Forks
- 16
- Métricas de merge de PR
- Sin PR fusionados en 30 d
Descripción
## Background
Right now, based on #1538, the DataFrame is re-read between each phase. This is potentially ineffective and results it extra IO (possible extra cost in cloud)
## Feature
Investigate and eventually implement some Spark caching strategy resistant to node failure.
## Proposed Solution
This might be a source reading for possible solution
https://sparkbyexamples.com/spark/spark-dataframe-cache-and-persist-explained/
Guía de contribución
Evaluación
Este issue todavía no se ha evaluado.