Initial Spark Prototype Development

Matei Zaharia begins developing the first working prototype of Spark, driven by the limitations of Hadoop MapReduce for iterative applications and interactive queries.

Significance

This marked the birth of Spark, laying the technical foundations for in-memory, fault-tolerant distributed processing and a new paradigm for big data analytics.

Key facts

Year
2009
Type
invention
Location
UC Berkeley AMPLab, Berkeley, California, USA