Dremel: A Decade of Interactive Big-Data Analysis at Google

Google publishes a seminal paper detailing the architecture and performance of Dremel, its internal interactive query system, which processes petabytes of data in seconds. This paper revealed the technological core that would become BigQuery.

Significance

This paper lifted the veil on Google's internal analytics prowess, showcasing a system far ahead of its time. It not only detailed the column-oriented storage, nested data types, and tree-based execution model but also ignited discussions and inspired similar systems (like Apache Drill, Impala, Presto) in the open-source community.

Key facts

Date
2010-09-01
Type
publication
Location
ACM SIGMOD International Conference on Management of Data