Re:PipelineDB v0.9.9 – One More Release Until PipelineDB Is ...
by 75dvtwin on Wednesday April 25, 2018 @04:59PM
(#16926085)
Pipeline DB would be a great fit, it seems, for 'Speed-layer' part of lambda architecture. While the postgresql underneath would work as a 'batch layer'.
A pair of both would only be missing an in-memory and on disk column-oriented 'analytics' tier (perhaps as separate instance that's being continuously fed from the first two?).
I think I had seen that PostgreSQL also has columnar store extension, wondering if that would work close the remaining gap for the analytics instance.
Re:PipelineDB v0.9.9 – One More Release Until PipelineDB Is ...
by chucky_z on Wednesday April 25, 2018 @07:46PM
(#16927517)
fdw_cstore! It's a workable columnar store, but honestly if you have "big" data it's only passable. It is much easier to run than other columar stores however.
A pair of both would only be missing an in-memory and on disk column-oriented 'analytics' tier (perhaps as separate instance that's being continuously fed from the first two?).
I think I had seen that PostgreSQL also has columnar store extension, wondering if that would work close the remaining gap for the analytics instance.