> I've just uploaded ZSON extension on GitHub: > > https://github.com/afiskon/zson > > ZSON learns on your common JSONB documents and creates a dictionary > with strings that are frequently used in all documents. After that you > can use ZSON type to compress documents using this dictionary. When > documents schema evolve and compression becomes inefficient you can > re-learn on new documents. New documents will be compressed with a new > dictionary, old documents will be decompressed using old dictionary. > > In some cases ZSON can save half of your disk space and give you about > 10% more TPS. Everything depends on your data and workload though. > Memory is saved as well. For more details see README.md. > > Please don't hesitate to ask any questions. Any feedback and pull > requests are welcome too!
Very good. Oleg had mentioned that dictionary compression was being considered.
It would be useful to be able to define compression dictionaries for many use cases.
Dictionary compression is a different project, we'll publish it after testing.
Will you be submitting this to core? -- Simon Riggs http://www.2ndQuadrant.com/ PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services