MapReduce, Hadoop and Spark (big-data processing)
Programming model for parallel batch jobs over clusters of cheap machines; Hadoop open-sourced it and launched the 'big data' industry.
- 2004

First version
Jeffrey Dean and Sanjay Ghemawat (Google); Hadoop by Doug Cutting and Mike Cafarella; Spark by Matei Zaharia (UC Berkeley)
Image: Google Inc. (logo by Ruth Kedar, 1999) · Unclear; no licence stated (Google trad… · google.com home page logo file, Wayback Machine… - 2006

Apache Hadoop
Open-source MapReduce and HDFS, scaled at Yahoo, spreads big-data clusters.
Image: Apache Software Foundation · Apache License 2.0 · Wikimedia Commons - 2010

Spark research release
In-memory resilient distributed datasets make iterative analytics far faster than MapReduce.
Image: Apache Software Foundation · Apache License 2.0 · Wikimedia Commons - 2011

Apache Kafka stream processing
LinkedIn open-sources a distributed log that moves big-data work from nightly batches to continuous real-time streams.
Image: ™/®Apache Software Foundation · Apache License 2.0 · Wikimedia Commons