From LucidDB Wiki
Revision as of 18:29, 14 August 2009 by Rchen
This page links to some published benchmarks and applications which could serve as good showcases for LucidDB's architecture.
- TPC-H: the load and single-stream query portion of this has been known to run successfully at up to 30 gigabyte scale, with the omission/modification of a few queries where subquery optimization still needs work. The concurrent-stream query portion should be able to run too. The concurrent read/write portion can't be implemented exactly as specified since LucidDB does not yet support user-level transactions; a modified version with autocommit should be able to work. Details at LucidDbTpch.
- DBT-3 is a fair usage implementation of the TPC-H specifications.
- TPC-R: based on the same schema as TPC-H. Marked as obsolete in 2005, but may still produce interesting results.
- TPC-DS: currently under development; a draft version is available.
- SSB: a simplified version of TPC-H which eliminates non-star structure. (Download dbgen source here)
- DWEB: a highly synthetic meta-benchmark which can be used to generate various physical schemas (star, snowflake, constellation) from the same logical schema or subsets thereof