"Big data" features coming in PostgreSQL 9.5
"Big data" features coming in PostgreSQL 9.5
Posted Aug 5, 2015 20:13 UTC (Wed) by xtifr (guest, #143)Parent article: "Big data" features coming in PostgreSQL 9.5
BRIN indexes are niche! It's just a fairly common and very important niche. :)
Faster indexing and, to a lesser extent, BRIN both sound like features that will seriously benefit most users, not just Big Data.
PostgreSQL used to have the reputation of being the slow-but-reliable FOSS dbms. The reliable part may have taken a minor ding with the recent multixact problems, but I think they can get over that. I think the slow part is already much less true than it used to be, but I still approve of any and all speedups they can add. If they became the faster-and-more-reliable FOSS dbms, I wouldn't mind one bit!
(I really hope the planned Parallel Table Scan thing works out well, and provides a decent performance boost as well.)
Posted Aug 5, 2015 20:56 UTC (Wed)
by kleptog (subscriber, #1183)
[Link]
Yes! One thing is saw in the initial discussions but I'm not seeing mentioned anywhere is that the theory behind BRIN indexes should be able to be applied to support Bloom filters. And they are seriously useful, especially for high cardinality columns which sort really badly. Btrees don't do nicely on those kind of columns. I'm hoping that they haven't been forgotten.
Posted Aug 6, 2015 22:52 UTC (Thu)
by xtifr (guest, #143)
[Link]
"Big data" features coming in PostgreSQL 9.5
Er, I meant to say, "faster sorting" sounds like a feature for everyone, not, "faster indexing". In case it wasn't obvious. Had indexes on the brain.
"Big data" features coming in PostgreSQL 9.5