Unanswered: Schema Evolution Research: willing to participate?
I'm Carlo Curino a researcher at UCLA and Politecnico di Milano.
I work on Schema Evolution and Transaction Time databases.
I'm working on creating a Benchmark for Schema Evolution based on REAL systems. I started by dissecting the evolution of MediaWiki (the software platform of Wikipedia) and its 170+ versions of the schema (MySQL DB backend). You can start using the data I collected including the REAL wikipedia query templates: http://yellowstone.cs.ucla.edu/schem...tion_Benchmark
I'm extending it with other open-source Web Information Systems (TikiWiki, Joomla, Slashcode, Coopermine, Typo3, Xoops, etc...).
I would like to further increase the pool of systems to be included in the Benchmark with commercial (traditional) DB.
If any of you is willing to cooperate to this, we will cite and acknowledge you and your company in our websites and in the final Benchmark documentation.
Your role would be collecting (or let me collect):
- the history of schema versions (10+)
- SQL update scripts modifying the schema and migrating the data (if available)
- a set of template of queries (better under each schema version)
NO database content or privacy critical data are required nor accepted (e.g., do not release queries/updates containing actual values).
The goal is to develop a unified benchmark to test the various systems supporting schema evolution to provide a good way of comparing various solutions and to develop systems to support DBAs in the dirty job of Schema Evolution (see Screencast of our tool: http://yellowstone.cs.ucla.edu/schem...rism-Demo.mov).
Please contact me if interested
Last edited by carlocurino; 05-25-08 at 07:28.
Reason: layout correction