Clutch DB - Get in gear

What is ClutchDB

ClutchDB is a distributed SQL compute cluster for MySQL compatible databases (MySQL/MariaDB/Percona Server) with a focus on OLAP workloads. ClutchDB is designed to be deployed in the cloud. The name comes from the word 'clutch' as in a 'a clutch of eggs'.

Many of the components are fully developed as stand-alone projects and the ClutchDB simply represents integration of the components.

Improved "in the cloud" or "in the datacenter" performance

• Scale Up - Adds parallelism for many types of queries
• Scale Out - Spread data across multiple nodes for linear scalability or better

What kind of partitioning can I use

• Directory based partitioning (one directory server, many storage servers)
• Hash based partitioning

Can I modify data?

• Bulk loading - uses named pipes and LOAD DATA LOCAL INFILE to load tables
• The bundled storage engine (Infobright Community Edition) only supports LOAD DATA for adding data

Future Roadmap

• Range based partitioning
• Materialized views
• Bitmap indexes (not storage engine dependent)
• Star schema optimization (with bitmap indexes)
• Easy node splitting (not supported with hash partitions)