On 29 July 2016 at 15:44, Stephen Frost <sfrost@snowman.net> wrote:
> All very hand-wavy, of course, and it'd make sense to make the concept
> work for BRIN before we consider anything else, but it seems like there
> could be a use-case for allowing indexes other than BRIN to be built in
> a way that allows HOT updates to happen, thus eliminating the cost of
> having to update those indexes when the tuple is changed, in many cases.
> Of course, those indexes couldn't be used UNIQUE indexes or used for
> primary keys, and adjusting the parameters to a BRIN index you could
> possibly get a similar index, but this might allow such an index to
> still be usable for index-only scans, which a BRIN index will never be
> able to provide.
This idea is vaguely similar to the concepts I've been working on,
based on earlier work in 2007.
I'm starting a new post for a full discussion.
--
Simon Riggs http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services