4 ms·
(Ozgun from Citus Data) Sure, we'd be interested in running more numbers. If you have example data sets in mind, could you share them with us? For clarificati
by ozgune 11y ago
(Ozgun from Citus Data)
Sure, we'd be interested in running more numbers. If you have example data sets in mind, could you share them with us?
For clarification, we picked this data set for several reasons. The data set was real, sizeable, publicly available, and it became highly referenced in PostgreSQL's JSON/JSONB development:
http://www.pgcon.org/2014/schedule/attachments/328_9.4json.pdf http://www.pgcon.org/2014/schedule/attachments/328_9.4json.p...
http://www.pgcon.org/2014/schedule/attachments/313_xml-hstore-json.pdf http://www.pgcon.org/2014/schedule/attachments/313_xml-hstor...
http://blog.2ndquadrant.com/jsonb-type-performance-postgresql-9-4/ http://blog.2ndquadrant.com/jsonb-type-performance-postgresq...
http://www.pgcon.org/2014/schedule/attachments/318_pgcon-2014-vodka.pdf http://www.pgcon.org/2014/schedule/attachments/318_pgcon-201...
- ddorian43 11y agoCan you explain HOW are you compressing the json data? Ex, is it just block-pgzip-compress? Or are you exploding each jsonb-field as a separate file like with normal columns ?