This talk will cover Pilosa - a high performance, distributed bitmap index written in Go. Bitmap indexes are widely used inside various data systems to power difficult analytical queries; while conceptually simple, research and improvements into storing and computing over bitmaps has continued unabated for decades. Pilosa leverages the latest and most successful bitmap techniques to power a distributed data indexing system that excels at a variety of analytical workloads while consuming less space and memory than many similar systems. We'll take a look at how Pilosa uses distributed bitmaps to support multi-field GROUP BY queries and very complex WHERE clauses over billions of records with sub-second query performance. Perhaps more importantly, we'll discuss how Go has enabled us to productively develop this technology and deploy it in various production systems.
Unfortunately, the view of the audience was recorded rather than the slides and speaker. The slides however are available here: